diff --git a/README.md b/README.md
index ed10a950..e36ae18a 100644
--- a/README.md
+++ b/README.md
@@ -23,8 +23,8 @@ details](https://github.com/ssube/onnx-web/blob/main/docs/user-guide.md).
 
 This is an incomplete list of new and interesting features, with links to the user guide:
 
-- SDXL support
-- LCM support
+- supports SDXL and SDXL Turbo
+- wide variety of schedulers: DDIM, DEIS, DPM SDE, Euler Ancestral, LCM, UniPC, and more
 - hardware acceleration on both AMD and Nvidia
   - tested on CUDA, DirectML, and ROCm
   - [half-precision support for low-memory GPUs](docs/user-guide.md#optimizing-models-for-lower-memory-usage) on both
diff --git a/api/launch.bat b/api/launch.bat
index 0d5933e5..680de5d9 100644
--- a/api/launch.bat
+++ b/api/launch.bat
@@ -1,5 +1,7 @@
 call onnx_env\Scripts\Activate.bat
 
+echo "This launch.bat script is deprecated in favor of launch.ps1 and will be removed in a future release."
+
 echo "Downloading and converting models to ONNX format..."
 IF "%ONNX_WEB_EXTRA_MODELS%"=="" (set ONNX_WEB_EXTRA_MODELS=..\models\extras.json)
 python -m onnx_web.convert ^
@@ -10,6 +12,12 @@ python -m onnx_web.convert ^
 --extras=%ONNX_WEB_EXTRA_MODELS% ^
 --token=%HF_TOKEN% %ONNX_WEB_EXTRA_ARGS%
 
+IF NOT EXIST .\gui\index.html (
+  echo "Please make sure you have downloaded the web UI files from https://github.com/ssube/onnx-web/tree/gh-pages"
+  echo "See https://github.com/ssube/onnx-web/blob/main/docs/setup-guide.md#download-the-web-ui-bundle for more information"
+  pause
+)
+
 echo "Launching API server..."
 waitress-serve ^
 --host=0.0.0.0 ^
diff --git a/api/launch.ps1 b/api/launch.ps1
index 19add137..ba898151 100644
--- a/api/launch.ps1
+++ b/api/launch.ps1
@@ -10,6 +10,13 @@ python -m onnx_web.convert `
 --extras=$Env:ONNX_WEB_EXTRA_MODELS `
 --token=$Env:HF_TOKEN $Env:ONNX_WEB_EXTRA_ARGS
 
+if (!(Test-Path -path .\gui\index.html -PathType Leaf)) {
+  echo "Downloading latest web UI files from Github..."
+  Invoke-WebRequest "https://raw.githubusercontent.com/ssube/onnx-web/gh-pages/v0.11.0/index.html" -OutFile .\gui\index.html
+  Invoke-WebRequest "https://raw.githubusercontent.com/ssube/onnx-web/gh-pages/v0.11.0/config.json" -OutFile .\gui\config.json
+  Invoke-WebRequest "https://raw.githubusercontent.com/ssube/onnx-web/gh-pages/v0.11.0/bundle/main.js" -OutFile .\gui\bundle\main.js
+}
+
 echo "Launching API server..."
 waitress-serve `
 --host=0.0.0.0 `
diff --git a/api/onnx_web/chain/blend_mask.py b/api/onnx_web/chain/blend_mask.py
index 4486bbf6..326ba162 100644
--- a/api/onnx_web/chain/blend_mask.py
+++ b/api/onnx_web/chain/blend_mask.py
@@ -30,17 +30,16 @@ class BlendMaskStage(BaseStage):
     ) -> StageResult:
         logger.info("blending image using mask")
 
-        # TODO: does this need an alpha channel?
         mult_mask = Image.new(stage_mask.mode, stage_mask.size, color="black")
-        mult_mask.alpha_composite(stage_mask)
+        mult_mask = Image.alpha_composite(mult_mask, stage_mask)
         mult_mask = mult_mask.convert("L")
 
         if is_debug():
             save_image(server, "last-mask.png", stage_mask)
             save_image(server, "last-mult-mask.png", mult_mask)
 
-        return StageResult(
-            images=[
+        return StageResult.from_images(
+            [
                 Image.composite(stage_source, source, mult_mask)
                 for source in sources.as_image()
             ]
diff --git a/api/onnx_web/convert/__main__.py b/api/onnx_web/convert/__main__.py
index 5cbe7f07..1a4f0241 100644
--- a/api/onnx_web/convert/__main__.py
+++ b/api/onnx_web/convert/__main__.py
@@ -3,16 +3,17 @@ from argparse import ArgumentParser
 from logging import getLogger
 from os import makedirs, path
 from sys import exit
-from typing import Any, Dict, List, Optional, Tuple
-from urllib.parse import urlparse
+from typing import Any, Dict, List, Union
 
-from huggingface_hub.file_download import hf_hub_download
 from jsonschema import ValidationError, validate
 from onnx import load_model, save_model
 from transformers import CLIPTokenizer
 
 from ..constants import ONNX_MODEL, ONNX_WEIGHTS
+from ..server.plugin import load_plugins
 from ..utils import load_config
+from .client import add_model_source, fetch_model
+from .client.huggingface import HuggingfaceClient
 from .correction.gfpgan import convert_correction_gfpgan
 from .diffusion.control import convert_diffusion_control
 from .diffusion.diffusion import convert_diffusion_diffusers
@@ -25,8 +26,7 @@ from .upscaling.swinir import convert_upscaling_swinir
 from .utils import (
     DEFAULT_OPSET,
     ConversionContext,
-    download_progress,
-    remove_prefix,
+    fix_diffusion_name,
     source_format,
     tuple_to_correction,
     tuple_to_diffusion,
@@ -44,32 +44,34 @@ warnings.filterwarnings(
     ".*Converting a tensor to a Python boolean might cause the trace to be incorrect.*",
 )
 
-Models = Dict[str, List[Any]]
-
 logger = getLogger(__name__)
 
+ModelDict = Dict[str, Union[float, int, str]]
+Models = Dict[str, List[ModelDict]]
 
-model_sources: Dict[str, Tuple[str, str]] = {
-    "civitai://": ("Civitai", "https://civitai.com/api/download/models/%s"),
+model_converters: Dict[str, Any] = {
+    "img2img": convert_diffusion_diffusers,
+    "img2img-sdxl": convert_diffusion_diffusers_xl,
+    "inpaint": convert_diffusion_diffusers,
+    "txt2img": convert_diffusion_diffusers,
+    "txt2img-sdxl": convert_diffusion_diffusers_xl,
 }
 
-model_source_huggingface = "huggingface://"
-
 # recommended models
 base_models: Models = {
     "diffusion": [
         # v1.x
         (
             "stable-diffusion-onnx-v1-5",
-            model_source_huggingface + "runwayml/stable-diffusion-v1-5",
+            HuggingfaceClient.protocol + "runwayml/stable-diffusion-v1-5",
         ),
         (
             "stable-diffusion-onnx-v1-inpainting",
-            model_source_huggingface + "runwayml/stable-diffusion-inpainting",
+            HuggingfaceClient.protocol + "runwayml/stable-diffusion-inpainting",
         ),
         (
             "upscaling-stable-diffusion-x4",
-            model_source_huggingface + "stabilityai/stable-diffusion-x4-upscaler",
+            HuggingfaceClient.protocol + "stabilityai/stable-diffusion-x4-upscaler",
             True,
         ),
     ],
@@ -200,69 +202,203 @@ base_models: Models = {
 }
 
 
-def fetch_model(
-    conversion: ConversionContext,
-    name: str,
-    source: str,
-    dest: Optional[str] = None,
-    format: Optional[str] = None,
-    hf_hub_fetch: bool = False,
-    hf_hub_filename: Optional[str] = None,
-) -> Tuple[str, bool]:
-    cache_path = dest or conversion.cache_path
-    cache_name = path.join(cache_path, name)
+def convert_model_source(conversion: ConversionContext, model):
+    model_format = source_format(model)
+    name = model["name"]
+    source = model["source"]
 
-    # add an extension if possible, some of the conversion code checks for it
-    if format is None:
-        url = urlparse(source)
-        ext = path.basename(url.path)
-        _filename, ext = path.splitext(ext)
-        if ext is not None:
-            cache_name = cache_name + ext
+    dest_path = None
+    if "dest" in model:
+        dest_path = path.join(conversion.model_path, model["dest"])
+
+    dest = fetch_model(conversion, name, source, format=model_format, dest=dest_path)
+    logger.info("finished downloading source: %s -> %s", source, dest)
+
+
+def convert_model_network(conversion: ConversionContext, model):
+    model_format = source_format(model)
+    model_type = model["type"]
+    name = model["name"]
+    source = model["source"]
+
+    if model_type == "control":
+        dest = fetch_model(
+            conversion,
+            name,
+            source,
+            format=model_format,
+        )
+
+        convert_diffusion_control(
+            conversion,
+            model,
+            dest,
+            path.join(conversion.model_path, model_type, name),
+        )
     else:
-        cache_name = f"{cache_name}.{format}"
+        model = model.get("model", None)
+        dest = fetch_model(
+            conversion,
+            name,
+            source,
+            dest=path.join(conversion.model_path, model_type),
+            format=model_format,
+            embeds=(model_type == "inversion" and model == "concept"),
+        )
 
-    if path.exists(cache_name):
-        logger.debug("model already exists in cache, skipping fetch")
-        return cache_name, False
+    logger.info("finished downloading network: %s -> %s", source, dest)
 
-    for proto in model_sources:
-        api_name, api_root = model_sources.get(proto)
-        if source.startswith(proto):
-            api_source = api_root % (remove_prefix(source, proto))
-            logger.info(
-                "downloading model from %s: %s -> %s", api_name, api_source, cache_name
+
+def convert_model_diffusion(conversion: ConversionContext, model):
+    # fix up entries with missing prefixes
+    name = fix_diffusion_name(model["name"])
+    if name != model["name"]:
+        # update the model in-memory if the name changed
+        model["name"] = name
+
+    model_format = source_format(model)
+
+    pipeline = model.get("pipeline", "txt2img")
+    converter = model_converters.get(pipeline)
+    converted, dest = converter(
+        conversion,
+        model,
+        model_format,
+    )
+
+    # make sure blending only happens once, not every run
+    if converted:
+        # keep track of which models have been blended
+        blend_models = {}
+
+        inversion_dest = path.join(conversion.model_path, "inversion")
+        lora_dest = path.join(conversion.model_path, "lora")
+
+        for inversion in model.get("inversions", []):
+            if "text_encoder" not in blend_models:
+                blend_models["text_encoder"] = load_model(
+                    path.join(
+                        dest,
+                        "text_encoder",
+                        ONNX_MODEL,
+                    )
+                )
+
+            if "tokenizer" not in blend_models:
+                blend_models["tokenizer"] = CLIPTokenizer.from_pretrained(
+                    dest,
+                    subfolder="tokenizer",
+                )
+
+            inversion_name = inversion["name"]
+            inversion_source = inversion["source"]
+            inversion_format = inversion.get("format", None)
+            inversion_source = fetch_model(
+                conversion,
+                inversion_name,
+                inversion_source,
+                dest=inversion_dest,
             )
-            return download_progress([(api_source, cache_name)]), False
+            inversion_token = inversion.get("token", inversion_name)
+            inversion_weight = inversion.get("weight", 1.0)
 
-    if source.startswith(model_source_huggingface):
-        hub_source = remove_prefix(source, model_source_huggingface)
-        logger.info("downloading model from Huggingface Hub: %s", hub_source)
-        # from_pretrained has a bunch of useful logic that snapshot_download by itself down not
-        if hf_hub_fetch:
-            return (
-                hf_hub_download(
-                    repo_id=hub_source,
-                    filename=hf_hub_filename,
-                    cache_dir=cache_path,
-                    force_filename=f"{name}.bin",
-                ),
-                False,
+            blend_textual_inversions(
+                conversion,
+                blend_models["text_encoder"],
+                blend_models["tokenizer"],
+                [
+                    (
+                        inversion_source,
+                        inversion_weight,
+                        inversion_token,
+                        inversion_format,
+                    )
+                ],
             )
-        else:
-            return hub_source, True
-    elif source.startswith("https://"):
-        logger.info("downloading model from: %s", source)
-        return download_progress([(source, cache_name)]), False
-    elif source.startswith("http://"):
-        logger.warning("downloading model from insecure source: %s", source)
-        return download_progress([(source, cache_name)]), False
-    elif source.startswith(path.sep) or source.startswith("."):
-        logger.info("using local model: %s", source)
-        return source, False
+
+        for lora in model.get("loras", []):
+            if "text_encoder" not in blend_models:
+                blend_models["text_encoder"] = load_model(
+                    path.join(
+                        dest,
+                        "text_encoder",
+                        ONNX_MODEL,
+                    )
+                )
+
+            if "unet" not in blend_models:
+                blend_models["unet"] = load_model(path.join(dest, "unet", ONNX_MODEL))
+
+            # load models if not loaded yet
+            lora_name = lora["name"]
+            lora_source = lora["source"]
+            lora_source = fetch_model(
+                conversion,
+                f"{name}-lora-{lora_name}",
+                lora_source,
+                dest=lora_dest,
+            )
+            lora_weight = lora.get("weight", 1.0)
+
+            blend_loras(
+                conversion,
+                blend_models["text_encoder"],
+                [(lora_source, lora_weight)],
+                "text_encoder",
+            )
+
+            blend_loras(
+                conversion,
+                blend_models["unet"],
+                [(lora_source, lora_weight)],
+                "unet",
+            )
+
+        if "tokenizer" in blend_models:
+            dest_path = path.join(dest, "tokenizer")
+            logger.debug("saving blended tokenizer to %s", dest_path)
+            blend_models["tokenizer"].save_pretrained(dest_path)
+
+        for name in ["text_encoder", "unet"]:
+            if name in blend_models:
+                dest_path = path.join(dest, name, ONNX_MODEL)
+                logger.debug("saving blended %s model to %s", name, dest_path)
+                save_model(
+                    blend_models[name],
+                    dest_path,
+                    save_as_external_data=True,
+                    all_tensors_to_one_file=True,
+                    location=ONNX_WEIGHTS,
+                )
+
+
+def convert_model_upscaling(conversion: ConversionContext, model):
+    model_format = source_format(model)
+    name = model["name"]
+
+    source = fetch_model(conversion, name, model["source"], format=model_format)
+    model_type = model.get("model", "resrgan")
+    if model_type == "bsrgan":
+        convert_upscaling_bsrgan(conversion, model, source)
+    elif model_type == "resrgan":
+        convert_upscale_resrgan(conversion, model, source)
+    elif model_type == "swinir":
+        convert_upscaling_swinir(conversion, model, source)
     else:
-        logger.info("unknown model location, using path as provided: %s", source)
-        return source, False
+        logger.error("unknown upscaling model type %s for %s", model_type, name)
+        raise ValueError(name)
+
+
+def convert_model_correction(conversion: ConversionContext, model):
+    model_format = source_format(model)
+    name = model["name"]
+    source = fetch_model(conversion, name, model["source"], format=model_format)
+    model_type = model.get("model", "gfpgan")
+    if model_type == "gfpgan":
+        convert_correction_gfpgan(conversion, model, source)
+    else:
+        logger.error("unknown correction model type %s for %s", model_type, name)
+        raise ValueError(name)
 
 
 def convert_models(conversion: ConversionContext, args, models: Models):
@@ -276,69 +412,21 @@ def convert_models(conversion: ConversionContext, args, models: Models):
             if name in args.skip:
                 logger.info("skipping source: %s", name)
             else:
-                model_format = source_format(model)
-                source = model["source"]
-
                 try:
-                    dest_path = None
-                    if "dest" in model:
-                        dest_path = path.join(conversion.model_path, model["dest"])
-
-                    dest, hf = fetch_model(
-                        conversion, name, source, format=model_format, dest=dest_path
-                    )
-                    logger.info("finished downloading source: %s -> %s", source, dest)
+                    convert_model_source(conversion, model)
                 except Exception:
                     logger.exception("error fetching source %s", name)
                     model_errors.append(name)
 
     if args.networks and "networks" in models:
-        for network in models.get("networks", []):
-            name = network["name"]
+        for model in models.get("networks", []):
+            name = model["name"]
 
             if name in args.skip:
                 logger.info("skipping network: %s", name)
             else:
-                network_format = source_format(network)
-                network_model = network.get("model", None)
-                network_type = network["type"]
-                source = network["source"]
-
                 try:
-                    if network_type == "control":
-                        dest, hf = fetch_model(
-                            conversion,
-                            name,
-                            source,
-                            format=network_format,
-                        )
-
-                        convert_diffusion_control(
-                            conversion,
-                            network,
-                            dest,
-                            path.join(conversion.model_path, network_type, name),
-                        )
-                    if network_type == "inversion" and network_model == "concept":
-                        dest, hf = fetch_model(
-                            conversion,
-                            name,
-                            source,
-                            dest=path.join(conversion.model_path, network_type),
-                            format=network_format,
-                            hf_hub_fetch=True,
-                            hf_hub_filename="learned_embeds.bin",
-                        )
-                    else:
-                        dest, hf = fetch_model(
-                            conversion,
-                            name,
-                            source,
-                            dest=path.join(conversion.model_path, network_type),
-                            format=network_format,
-                        )
-
-                    logger.info("finished downloading network: %s -> %s", source, dest)
+                    convert_model_network(conversion, model)
                 except Exception:
                     logger.exception("error fetching network %s", name)
                     model_errors.append(name)
@@ -351,142 +439,8 @@ def convert_models(conversion: ConversionContext, args, models: Models):
             if name in args.skip:
                 logger.info("skipping model: %s", name)
             else:
-                model_format = source_format(model)
-
                 try:
-                    source, hf = fetch_model(
-                        conversion, name, model["source"], format=model_format
-                    )
-
-                    pipeline = model.get("pipeline", "txt2img")
-                    if pipeline.endswith("-sdxl"):
-                        converted, dest = convert_diffusion_diffusers_xl(
-                            conversion,
-                            model,
-                            source,
-                            model_format,
-                            hf=hf,
-                        )
-                    else:
-                        converted, dest = convert_diffusion_diffusers(
-                            conversion,
-                            model,
-                            source,
-                            model_format,
-                            hf=hf,
-                        )
-
-                    # make sure blending only happens once, not every run
-                    if converted:
-                        # keep track of which models have been blended
-                        blend_models = {}
-
-                        inversion_dest = path.join(conversion.model_path, "inversion")
-                        lora_dest = path.join(conversion.model_path, "lora")
-
-                        for inversion in model.get("inversions", []):
-                            if "text_encoder" not in blend_models:
-                                blend_models["text_encoder"] = load_model(
-                                    path.join(
-                                        dest,
-                                        "text_encoder",
-                                        ONNX_MODEL,
-                                    )
-                                )
-
-                            if "tokenizer" not in blend_models:
-                                blend_models[
-                                    "tokenizer"
-                                ] = CLIPTokenizer.from_pretrained(
-                                    dest,
-                                    subfolder="tokenizer",
-                                )
-
-                            inversion_name = inversion["name"]
-                            inversion_source = inversion["source"]
-                            inversion_format = inversion.get("format", None)
-                            inversion_source, hf = fetch_model(
-                                conversion,
-                                inversion_name,
-                                inversion_source,
-                                dest=inversion_dest,
-                            )
-                            inversion_token = inversion.get("token", inversion_name)
-                            inversion_weight = inversion.get("weight", 1.0)
-
-                            blend_textual_inversions(
-                                conversion,
-                                blend_models["text_encoder"],
-                                blend_models["tokenizer"],
-                                [
-                                    (
-                                        inversion_source,
-                                        inversion_weight,
-                                        inversion_token,
-                                        inversion_format,
-                                    )
-                                ],
-                            )
-
-                        for lora in model.get("loras", []):
-                            if "text_encoder" not in blend_models:
-                                blend_models["text_encoder"] = load_model(
-                                    path.join(
-                                        dest,
-                                        "text_encoder",
-                                        ONNX_MODEL,
-                                    )
-                                )
-
-                            if "unet" not in blend_models:
-                                blend_models["unet"] = load_model(
-                                    path.join(dest, "unet", ONNX_MODEL)
-                                )
-
-                            # load models if not loaded yet
-                            lora_name = lora["name"]
-                            lora_source = lora["source"]
-                            lora_source, hf = fetch_model(
-                                conversion,
-                                f"{name}-lora-{lora_name}",
-                                lora_source,
-                                dest=lora_dest,
-                            )
-                            lora_weight = lora.get("weight", 1.0)
-
-                            blend_loras(
-                                conversion,
-                                blend_models["text_encoder"],
-                                [(lora_source, lora_weight)],
-                                "text_encoder",
-                            )
-
-                            blend_loras(
-                                conversion,
-                                blend_models["unet"],
-                                [(lora_source, lora_weight)],
-                                "unet",
-                            )
-
-                        if "tokenizer" in blend_models:
-                            dest_path = path.join(dest, "tokenizer")
-                            logger.debug("saving blended tokenizer to %s", dest_path)
-                            blend_models["tokenizer"].save_pretrained(dest_path)
-
-                        for name in ["text_encoder", "unet"]:
-                            if name in blend_models:
-                                dest_path = path.join(dest, name, ONNX_MODEL)
-                                logger.debug(
-                                    "saving blended %s model to %s", name, dest_path
-                                )
-                                save_model(
-                                    blend_models[name],
-                                    dest_path,
-                                    save_as_external_data=True,
-                                    all_tensors_to_one_file=True,
-                                    location=ONNX_WEIGHTS,
-                                )
-
+                    convert_model_diffusion(conversion, model)
                 except Exception:
                     logger.exception(
                         "error converting diffusion model %s",
@@ -502,24 +456,8 @@ def convert_models(conversion: ConversionContext, args, models: Models):
             if name in args.skip:
                 logger.info("skipping model: %s", name)
             else:
-                model_format = source_format(model)
-
                 try:
-                    source, hf = fetch_model(
-                        conversion, name, model["source"], format=model_format
-                    )
-                    model_type = model.get("model", "resrgan")
-                    if model_type == "bsrgan":
-                        convert_upscaling_bsrgan(conversion, model, source)
-                    elif model_type == "resrgan":
-                        convert_upscale_resrgan(conversion, model, source)
-                    elif model_type == "swinir":
-                        convert_upscaling_swinir(conversion, model, source)
-                    else:
-                        logger.error(
-                            "unknown upscaling model type %s for %s", model_type, name
-                        )
-                        model_errors.append(name)
+                    convert_model_upscaling(conversion, model)
                 except Exception:
                     logger.exception(
                         "error converting upscaling model %s",
@@ -535,19 +473,8 @@ def convert_models(conversion: ConversionContext, args, models: Models):
             if name in args.skip:
                 logger.info("skipping model: %s", name)
             else:
-                model_format = source_format(model)
                 try:
-                    source, hf = fetch_model(
-                        conversion, name, model["source"], format=model_format
-                    )
-                    model_type = model.get("model", "gfpgan")
-                    if model_type == "gfpgan":
-                        convert_correction_gfpgan(conversion, model, source)
-                    else:
-                        logger.error(
-                            "unknown correction model type %s for %s", model_type, name
-                        )
-                        model_errors.append(name)
+                    convert_model_correction(conversion, model)
                 except Exception:
                     logger.exception(
                         "error converting correction model %s",
@@ -559,12 +486,26 @@ def convert_models(conversion: ConversionContext, args, models: Models):
         logger.error("error while converting models: %s", model_errors)
 
 
+def register_plugins(conversion: ConversionContext):
+    logger.info("loading conversion plugins")
+    exports = load_plugins(conversion)
+
+    for proto, client in exports.clients:
+        try:
+            add_model_source(proto, client)
+        except Exception:
+            logger.exception("error loading client for protocol: %s", proto)
+
+    # TODO: add converters
+
+
 def main(args=None) -> int:
     parser = ArgumentParser(
         prog="onnx-web model converter", description="convert checkpoint models to ONNX"
     )
 
     # model groups
+    parser.add_argument("--base", action="store_true", default=True)
     parser.add_argument("--networks", action="store_true", default=True)
     parser.add_argument("--sources", action="store_true", default=True)
     parser.add_argument("--correction", action="store_true", default=False)
@@ -602,16 +543,20 @@ def main(args=None) -> int:
     server.half = args.half or server.has_optimization("onnx-fp16")
     server.opset = args.opset
     server.token = args.token
+
+    register_plugins(server)
+
     logger.info(
-        "converting models in %s using %s", server.model_path, server.training_device
+        "converting models into %s using %s", server.model_path, server.training_device
     )
 
     if not path.exists(server.model_path):
         logger.info("model path does not existing, creating: %s", server.model_path)
         makedirs(server.model_path)
 
-    logger.info("converting base models")
-    convert_models(server, args, base_models)
+    if args.base:
+        logger.info("converting base models")
+        convert_models(server, args, base_models)
 
     extras = []
     extras.extend(server.extra_models)
diff --git a/api/onnx_web/convert/client/__init__.py b/api/onnx_web/convert/client/__init__.py
new file mode 100644
index 00000000..1d90f04e
--- /dev/null
+++ b/api/onnx_web/convert/client/__init__.py
@@ -0,0 +1,54 @@
+from typing import Callable, Dict, Optional
+from logging import getLogger
+from os import path
+
+from ..utils import ConversionContext
+from .base import BaseClient
+from .civitai import CivitaiClient
+from .file import FileClient
+from .http import HttpClient
+from .huggingface import HuggingfaceClient
+
+logger = getLogger(__name__)
+
+
+model_sources: Dict[str, Callable[[], BaseClient]] = {
+    CivitaiClient.protocol: CivitaiClient,
+    FileClient.protocol: FileClient,
+    HttpClient.insecure_protocol: HttpClient,
+    HttpClient.protocol: HttpClient,
+    HuggingfaceClient.protocol: HuggingfaceClient,
+}
+
+
+def add_model_source(proto: str, client: BaseClient):
+    global model_sources
+
+    if proto in model_sources:
+        raise ValueError("protocol has already been taken")
+
+    model_sources[proto] = client
+
+
+def fetch_model(
+    conversion: ConversionContext,
+    name: str,
+    source: str,
+    format: Optional[str] = None,
+    dest: Optional[str] = None,
+    **kwargs,
+) -> str:
+    # TODO: switch to urlparse's default scheme
+    if source.startswith(path.sep) or source.startswith("."):
+        logger.info("adding file protocol to local path source: %s", source)
+        source = FileClient.protocol + source
+
+    for proto, client_type in model_sources.items():
+        if source.startswith(proto):
+            client = client_type(conversion)
+            return client.download(
+                conversion, name, source, format=format, dest=dest, **kwargs
+            )
+
+    logger.warning("unknown model protocol, using path as provided: %s", source)
+    return source
diff --git a/api/onnx_web/convert/client/base.py b/api/onnx_web/convert/client/base.py
new file mode 100644
index 00000000..2ec2ae65
--- /dev/null
+++ b/api/onnx_web/convert/client/base.py
@@ -0,0 +1,16 @@
+from typing import Optional
+
+from ..utils import ConversionContext
+
+
+class BaseClient:
+    def download(
+        self,
+        conversion: ConversionContext,
+        name: str,
+        source: str,
+        format: Optional[str] = None,
+        dest: Optional[str] = None,
+        **kwargs,
+    ) -> str:
+        raise NotImplementedError()
diff --git a/api/onnx_web/convert/client/civitai.py b/api/onnx_web/convert/client/civitai.py
new file mode 100644
index 00000000..b57b6533
--- /dev/null
+++ b/api/onnx_web/convert/client/civitai.py
@@ -0,0 +1,64 @@
+from logging import getLogger
+from typing import Optional
+
+from ..utils import (
+    ConversionContext,
+    build_cache_paths,
+    download_progress,
+    get_first_exists,
+    remove_prefix,
+)
+from .base import BaseClient
+
+logger = getLogger(__name__)
+
+CIVITAI_ROOT = "https://civitai.com/api/download/models/%s"
+
+
+class CivitaiClient(BaseClient):
+    name = "civitai"
+    protocol = "civitai://"
+
+    root: str
+    token: Optional[str]
+
+    def __init__(
+        self,
+        conversion: ConversionContext,
+        token: Optional[str] = None,
+        root: str = CIVITAI_ROOT,
+    ):
+        self.root = conversion.get_setting("CIVITAI_ROOT", root)
+        self.token = conversion.get_setting("CIVITAI_TOKEN", token)
+
+    def download(
+        self,
+        conversion: ConversionContext,
+        name: str,
+        source: str,
+        format: Optional[str] = None,
+        dest: Optional[str] = None,
+        **kwargs,
+    ) -> str:
+        cache_paths = build_cache_paths(
+            conversion,
+            name,
+            client=CivitaiClient.name,
+            format=format,
+            dest=dest,
+        )
+        cached = get_first_exists(cache_paths)
+        if cached:
+            return cached
+
+        source = self.root % (remove_prefix(source, CivitaiClient.protocol))
+        logger.info("downloading model from Civitai: %s -> %s", source, cache_paths[0])
+
+        if self.token:
+            logger.debug("adding Civitai token authentication")
+            if "?" in source:
+                source = f"{source}&token={self.token}"
+            else:
+                source = f"{source}?token={self.token}"
+
+        return download_progress(source, cache_paths[0])
diff --git a/api/onnx_web/convert/client/file.py b/api/onnx_web/convert/client/file.py
new file mode 100644
index 00000000..8ed1a4d9
--- /dev/null
+++ b/api/onnx_web/convert/client/file.py
@@ -0,0 +1,32 @@
+from logging import getLogger
+from os import path
+from typing import Optional
+from urllib.parse import urlparse
+
+from ..utils import ConversionContext
+from .base import BaseClient
+
+logger = getLogger(__name__)
+
+
+class FileClient(BaseClient):
+    protocol = "file://"
+
+    def __init__(self, _conversion: ConversionContext):
+        """
+        Nothing to initialize for this client.
+        """
+        pass
+
+    def download(
+        self,
+        conversion: ConversionContext,
+        _name: str,
+        uri: str,
+        format: Optional[str] = None,
+        dest: Optional[str] = None,
+        **kwargs,
+    ) -> str:
+        parts = urlparse(uri)
+        logger.info("loading model from: %s", parts.path)
+        return path.join(dest or conversion.model_path, parts.path)
diff --git a/api/onnx_web/convert/client/http.py b/api/onnx_web/convert/client/http.py
new file mode 100644
index 00000000..151ebccd
--- /dev/null
+++ b/api/onnx_web/convert/client/http.py
@@ -0,0 +1,52 @@
+from logging import getLogger
+from typing import Dict, Optional
+
+from ..utils import (
+    ConversionContext,
+    build_cache_paths,
+    download_progress,
+    get_first_exists,
+)
+from .base import BaseClient
+
+logger = getLogger(__name__)
+
+
+class HttpClient(BaseClient):
+    name = "http"
+    protocol = "https://"
+    insecure_protocol = "http://"
+
+    headers: Dict[str, str]
+
+    def __init__(
+        self, _conversion: ConversionContext, headers: Optional[Dict[str, str]] = None
+    ):
+        self.headers = headers or {}
+
+    def download(
+        self,
+        conversion: ConversionContext,
+        name: str,
+        source: str,
+        format: Optional[str] = None,
+        dest: Optional[str] = None,
+        **kwargs,
+    ) -> str:
+        cache_paths = build_cache_paths(
+            conversion,
+            name,
+            client=HttpClient.name,
+            format=format,
+            dest=dest,
+        )
+        cached = get_first_exists(cache_paths)
+        if cached:
+            return cached
+
+        if source.startswith(HttpClient.protocol):
+            logger.info("downloading model from: %s", source)
+        elif source.startswith(HttpClient.insecure_protocol):
+            logger.warning("downloading model from insecure source: %s", source)
+
+        return download_progress(source, cache_paths[0])
diff --git a/api/onnx_web/convert/client/huggingface.py b/api/onnx_web/convert/client/huggingface.py
new file mode 100644
index 00000000..c35698e7
--- /dev/null
+++ b/api/onnx_web/convert/client/huggingface.py
@@ -0,0 +1,43 @@
+from logging import getLogger
+from typing import Optional
+
+from huggingface_hub.file_download import hf_hub_download
+
+from ..utils import ConversionContext, remove_prefix
+from .base import BaseClient
+
+logger = getLogger(__name__)
+
+
+class HuggingfaceClient(BaseClient):
+    name = "huggingface"
+    protocol = "huggingface://"
+
+    token: Optional[str]
+
+    def __init__(self, conversion: ConversionContext, token: Optional[str] = None):
+        self.token = conversion.get_setting("HUGGINGFACE_TOKEN", token)
+
+    def download(
+        self,
+        conversion: ConversionContext,
+        name: str,
+        source: str,
+        format: Optional[str] = None,
+        dest: Optional[str] = None,
+        embeds: bool = False,
+        **kwargs,
+    ) -> str:
+        source = remove_prefix(source, HuggingfaceClient.protocol)
+        logger.info("downloading model from Huggingface Hub: %s", source)
+
+        if embeds:
+            return hf_hub_download(
+                repo_id=source,
+                filename="learned_embeds.bin",
+                cache_dir=dest or conversion.cache_path,
+                force_filename=f"{name}.bin",
+                token=self.token,
+            )
+        else:
+            return source
diff --git a/api/onnx_web/convert/diffusion/diffusion.py b/api/onnx_web/convert/diffusion/diffusion.py
index a8ecbbf7..a2e3d807 100644
--- a/api/onnx_web/convert/diffusion/diffusion.py
+++ b/api/onnx_web/convert/diffusion/diffusion.py
@@ -36,6 +36,8 @@ from ...diffusers.pipelines.upscale import OnnxStableDiffusionUpscalePipeline
 from ...diffusers.version_safe_diffusers import AttnProcessor
 from ...models.cnet import UNet2DConditionModel_CNet
 from ...utils import run_gc
+from ..client import fetch_model
+from ..client.huggingface import HuggingfaceClient
 from ..utils import (
     RESOLVE_FORMATS,
     ConversionContext,
@@ -43,6 +45,7 @@ from ..utils import (
     is_torch_2_0,
     load_tensor,
     onnx_export,
+    remove_prefix,
 )
 from .checkpoint import convert_extract_checkpoint
 
@@ -267,14 +270,13 @@ def collate_cnet(cnet_path):
 def convert_diffusion_diffusers(
     conversion: ConversionContext,
     model: Dict,
-    source: str,
     format: Optional[str],
-    hf: bool = False,
 ) -> Tuple[bool, str]:
     """
     From https://github.com/huggingface/diffusers/blob/main/scripts/convert_stable_diffusion_checkpoint_to_onnx.py
     """
-    name = model.get("name")
+    name = str(model.get("name")).strip()
+    source = model.get("source")
 
     # optional
     config = model.get("config", None)
@@ -320,9 +322,11 @@ def convert_diffusion_diffusers(
             logger.info("ONNX model already exists, skipping")
             return (False, dest_path)
 
+    cache_path = fetch_model(conversion, name, source, format=format)
+
     pipe_class = CONVERT_PIPELINES.get(pipe_type)
     v2, pipe_args = get_model_version(
-        source, conversion.map_location, size=image_size, version=version
+        cache_path, conversion.map_location, size=image_size, version=version
     )
 
     is_inpainting = False
@@ -334,57 +338,66 @@ def convert_diffusion_diffusers(
         pipe_args["from_safetensors"] = True
 
     torch_source = None
-    if path.exists(source) and path.isdir(source):
-        logger.debug("loading pipeline from diffusers directory: %s", source)
-        pipeline = pipe_class.from_pretrained(
-            source,
-            torch_dtype=dtype,
-            use_auth_token=conversion.token,
-        ).to(device)
-    elif path.exists(source) and path.isfile(source):
-        if conversion.extract:
-            logger.debug("extracting SD checkpoint to Torch models: %s", source)
-            torch_source = convert_extract_checkpoint(
-                conversion,
-                source,
-                f"{name}-torch",
-                is_inpainting=is_inpainting,
-                config_file=config,
-                vae_file=replace_vae,
-            )
-            logger.debug("loading pipeline from extracted checkpoint: %s", torch_source)
+    if path.exists(cache_path):
+        if path.isdir(cache_path):
+            logger.debug("loading pipeline from diffusers directory: %s", source)
             pipeline = pipe_class.from_pretrained(
-                torch_source,
+                cache_path,
                 torch_dtype=dtype,
+                use_auth_token=conversion.token,
             ).to(device)
-
-            # VAE replacement already happened during extraction, skip
-            replace_vae = None
         else:
-            logger.debug("loading pipeline from SD checkpoint: %s", source)
-            pipeline = download_from_original_stable_diffusion_ckpt(
-                source,
-                original_config_file=config_path,
-                pipeline_class=pipe_class,
-                **pipe_args,
-            ).to(device, torch_dtype=dtype)
-    elif hf:
-        logger.debug("downloading pretrained model from Huggingface hub: %s", source)
+            if conversion.extract:
+                logger.debug("extracting SD checkpoint to Torch models: %s", source)
+                torch_source = convert_extract_checkpoint(
+                    conversion,
+                    source,
+                    f"{name}-torch",
+                    is_inpainting=is_inpainting,
+                    config_file=config,
+                    vae_file=replace_vae,
+                )
+                logger.debug(
+                    "loading pipeline from extracted checkpoint: %s", torch_source
+                )
+                pipeline = pipe_class.from_pretrained(
+                    torch_source,
+                    torch_dtype=dtype,
+                ).to(device)
+
+                # VAE replacement already happened during extraction, skip
+                replace_vae = None
+            else:
+                logger.debug("loading pipeline from SD checkpoint: %s", source)
+                pipeline = download_from_original_stable_diffusion_ckpt(
+                    cache_path,
+                    original_config_file=config_path,
+                    pipeline_class=pipe_class,
+                    **pipe_args,
+                ).to(device, torch_dtype=dtype)
+    elif source.startswith(HuggingfaceClient.protocol):
+        hf_path = remove_prefix(source, HuggingfaceClient.protocol)
+        logger.debug("downloading pretrained model from Huggingface hub: %s", hf_path)
         pipeline = pipe_class.from_pretrained(
-            source,
+            hf_path,
             torch_dtype=dtype,
             use_auth_token=conversion.token,
         ).to(device)
     else:
-        logger.warning("pipeline source not found or not recognized: %s", source)
-        raise ValueError(f"pipeline source not found or not recognized: {source}")
+        logger.warning(
+            "pipeline source not found and protocol not recognized: %s", source
+        )
+        raise ValueError(
+            f"pipeline source not found and protocol not recognized: {source}"
+        )
 
     if replace_vae is not None:
         vae_path = path.join(conversion.model_path, replace_vae)
-        if check_ext(replace_vae, RESOLVE_FORMATS):
+        vae_file = check_ext(vae_path, RESOLVE_FORMATS)
+        if vae_file[0]:
             pipeline.vae = AutoencoderKL.from_single_file(vae_path)
         else:
-            pipeline.vae = AutoencoderKL.from_pretrained(vae_path)
+            pipeline.vae = AutoencoderKL.from_pretrained(replace_vae)
 
     if is_torch_2_0:
         pipeline.unet.set_attn_processor(AttnProcessor())
diff --git a/api/onnx_web/convert/diffusion/diffusion_xl.py b/api/onnx_web/convert/diffusion/diffusion_xl.py
index 7a03b700..c09b9440 100644
--- a/api/onnx_web/convert/diffusion/diffusion_xl.py
+++ b/api/onnx_web/convert/diffusion/diffusion_xl.py
@@ -10,6 +10,7 @@ from onnxruntime.transformers.float16 import convert_float_to_float16
 from optimum.exporters.onnx import main_export
 
 from ...constants import ONNX_MODEL
+from ..client import fetch_model
 from ..utils import RESOLVE_FORMATS, ConversionContext, check_ext
 
 logger = getLogger(__name__)
@@ -19,14 +20,13 @@ logger = getLogger(__name__)
 def convert_diffusion_diffusers_xl(
     conversion: ConversionContext,
     model: Dict,
-    source: str,
     format: Optional[str],
-    hf: bool = False,
 ) -> Tuple[bool, str]:
     """
     From https://github.com/huggingface/diffusers/blob/main/scripts/convert_stable_diffusion_checkpoint_to_onnx.py
     """
-    name = model.get("name")
+    name = str(model.get("name")).strip()
+    source = model.get("source")
     replace_vae = model.get("vae", None)
 
     device = conversion.training_device
@@ -52,24 +52,26 @@ def convert_diffusion_diffusers_xl(
 
         return (False, dest_path)
 
+    cache_path = fetch_model(conversion, name, model["source"], format=format)
     # safetensors -> diffusers directory with torch models
     temp_path = path.join(conversion.cache_path, f"{name}-torch")
 
     if format == "safetensors":
         pipeline = StableDiffusionXLPipeline.from_single_file(
-            source, use_safetensors=True
+            cache_path, use_safetensors=True
         )
     else:
-        pipeline = StableDiffusionXLPipeline.from_pretrained(source)
+        pipeline = StableDiffusionXLPipeline.from_pretrained(cache_path)
 
     if replace_vae is not None:
         vae_path = path.join(conversion.model_path, replace_vae)
-        if check_ext(replace_vae, RESOLVE_FORMATS):
+        vae_file = check_ext(vae_path, RESOLVE_FORMATS)
+        if vae_file[0]:
             logger.debug("loading VAE from single tensor file: %s", vae_path)
             pipeline.vae = AutoencoderKL.from_single_file(vae_path)
         else:
-            logger.debug("loading pretrained VAE from path: %s", vae_path)
-            pipeline.vae = AutoencoderKL.from_pretrained(vae_path)
+            logger.debug("loading pretrained VAE from path: %s", replace_vae)
+            pipeline.vae = AutoencoderKL.from_pretrained(replace_vae)
 
     if path.exists(temp_path):
         logger.debug("torch model already exists for %s: %s", source, temp_path)
diff --git a/api/onnx_web/convert/utils.py b/api/onnx_web/convert/utils.py
index ef44ba20..7785007b 100644
--- a/api/onnx_web/convert/utils.py
+++ b/api/onnx_web/convert/utils.py
@@ -31,6 +31,15 @@ ModelDict = Dict[str, Union[str, int]]
 LegacyModel = Tuple[str, str, Optional[bool], Optional[bool], Optional[int]]
 
 DEFAULT_OPSET = 14
+DIFFUSION_PREFIX = [
+    "diffusion-",
+    "diffusion/",
+    "diffusion\\",
+    "stable-diffusion-",
+    "upscaling-",  # SD upscaling
+]
+MODEL_FORMATS = ["onnx", "pth", "ckpt", "safetensors"]
+RESOLVE_FORMATS = ["safetensors", "ckpt", "pt", "pth", "bin"]
 
 
 class ConversionContext(ServerContext):
@@ -85,38 +94,37 @@ class ConversionContext(ServerContext):
         return torch.device(self.training_device)
 
 
-def download_progress(urls: List[Tuple[str, str]]):
-    for url, dest in urls:
-        dest_path = Path(dest).expanduser().resolve()
-        dest_path.parent.mkdir(parents=True, exist_ok=True)
-
-        if dest_path.exists():
-            logger.debug("destination already exists: %s", dest_path)
-            return str(dest_path.absolute())
-
-        req = requests.get(
-            url,
-            stream=True,
-            allow_redirects=True,
-            headers={
-                "User-Agent": "onnx-web-api",
-            },
-        )
-        if req.status_code != 200:
-            req.raise_for_status()  # Only works for 4xx errors, per SO answer
-            raise RequestException(
-                "request to %s failed with status code: %s" % (url, req.status_code)
-            )
-
-        total = int(req.headers.get("Content-Length", 0))
-        desc = "unknown" if total == 0 else ""
-        req.raw.read = partial(req.raw.read, decode_content=True)
-        with tqdm.wrapattr(req.raw, "read", total=total, desc=desc) as data:
-            with dest_path.open("wb") as f:
-                shutil.copyfileobj(data, f)
+def download_progress(source: str, dest: str):
+    dest_path = Path(dest).expanduser().resolve()
+    dest_path.parent.mkdir(parents=True, exist_ok=True)
 
+    if dest_path.exists():
+        logger.debug("destination already exists: %s", dest_path)
         return str(dest_path.absolute())
 
+    req = requests.get(
+        source,
+        stream=True,
+        allow_redirects=True,
+        headers={
+            "User-Agent": "onnx-web-api",
+        },
+    )
+    if req.status_code != 200:
+        req.raise_for_status()  # Only works for 4xx errors, per SO answer
+        raise RequestException(
+            "request to %s failed with status code: %s" % (source, req.status_code)
+        )
+
+    total = int(req.headers.get("Content-Length", 0))
+    desc = "unknown" if total == 0 else ""
+    req.raw.read = partial(req.raw.read, decode_content=True)
+    with tqdm.wrapattr(req.raw, "read", total=total, desc=desc) as data:
+        with dest_path.open("wb") as f:
+            shutil.copyfileobj(data, f)
+
+    return str(dest_path.absolute())
+
 
 def tuple_to_source(model: Union[ModelDict, LegacyModel]):
     if isinstance(model, list) or isinstance(model, tuple):
@@ -184,10 +192,6 @@ def tuple_to_upscaling(model: Union[ModelDict, LegacyModel]):
         return model
 
 
-MODEL_FORMATS = ["onnx", "pth", "ckpt", "safetensors"]
-RESOLVE_FORMATS = ["safetensors", "ckpt", "pt", "pth", "bin"]
-
-
 def check_ext(name: str, exts: List[str]) -> Tuple[bool, str]:
     _name, ext = path.splitext(name)
     ext = ext.strip(".")
@@ -215,6 +219,9 @@ def remove_prefix(name: str, prefix: str) -> str:
 
 
 def load_torch(name: str, map_location=None) -> Optional[Dict]:
+    """
+    TODO: move out of convert
+    """
     try:
         logger.debug("loading tensor with Torch: %s", name)
         checkpoint = torch.load(name, map_location=map_location)
@@ -228,6 +235,9 @@ def load_torch(name: str, map_location=None) -> Optional[Dict]:
 
 
 def load_tensor(name: str, map_location=None) -> Optional[Dict]:
+    """
+    TODO: move out of convert
+    """
     logger.debug("loading tensor: %s", name)
     _, extension = path.splitext(name)
     extension = extension[1:].lower()
@@ -285,6 +295,9 @@ def load_tensor(name: str, map_location=None) -> Optional[Dict]:
 
 
 def resolve_tensor(name: str) -> Optional[str]:
+    """
+    TODO: move out of convert
+    """
     logger.debug("searching for tensors with known extensions: %s", name)
     for next_extension in RESOLVE_FORMATS:
         next_name = f"{name}.{next_extension}"
@@ -345,3 +358,52 @@ def onnx_export(
             all_tensors_to_one_file=True,
             location=ONNX_WEIGHTS,
         )
+
+
+def fix_diffusion_name(name: str):
+    if not any([name.startswith(prefix) for prefix in DIFFUSION_PREFIX]):
+        logger.warning(
+            "diffusion models must have names starting with diffusion- to be recognized by the server: %s does not match",
+            name,
+        )
+        return f"diffusion-{name}"
+
+    return name
+
+
+def build_cache_paths(
+    conversion: ConversionContext,
+    name: str,
+    client: Optional[str] = None,
+    dest: Optional[str] = None,
+    format: Optional[str] = None,
+) -> List[str]:
+    cache_path = dest or conversion.cache_path
+
+    # add an extension if possible, some of the conversion code checks for it
+    if format is not None:
+        basename = path.basename(name)
+        _filename, ext = path.splitext(basename)
+        if ext is None or ext == "":
+            name = f"{name}.{format}"
+
+    paths = [
+        path.join(cache_path, name),
+    ]
+
+    if client is not None:
+        client_path = path.join(cache_path, client)
+        paths.append(path.join(client_path, name))
+
+    return paths
+
+
+def get_first_exists(
+    paths: List[str],
+) -> Optional[str]:
+    for name in paths:
+        if path.exists(name):
+            logger.debug("model already exists in cache, skipping fetch: %s", name)
+            return name
+
+    return None
diff --git a/api/onnx_web/diffusers/load.py b/api/onnx_web/diffusers/load.py
index 30a87863..b6620e8b 100644
--- a/api/onnx_web/diffusers/load.py
+++ b/api/onnx_web/diffusers/load.py
@@ -30,12 +30,12 @@ from .version_safe_diffusers import (
     DDPMScheduler,
     DEISMultistepScheduler,
     DPMSolverMultistepScheduler,
+    DPMSolverSDEScheduler,
     DPMSolverSinglestepScheduler,
     EulerAncestralDiscreteScheduler,
     EulerDiscreteScheduler,
     HeunDiscreteScheduler,
     IPNDMScheduler,
-    KarrasVeScheduler,
     KDPM2AncestralDiscreteScheduler,
     KDPM2DiscreteScheduler,
     LCMScheduler,
@@ -71,6 +71,7 @@ pipeline_schedulers = {
     "ddpm": DDPMScheduler,
     "deis-multi": DEISMultistepScheduler,
     "dpm-multi": DPMSolverMultistepScheduler,
+    "dpm-sde": DPMSolverSDEScheduler,
     "dpm-single": DPMSolverSinglestepScheduler,
     "euler": EulerDiscreteScheduler,
     "euler-a": EulerAncestralDiscreteScheduler,
@@ -78,7 +79,6 @@ pipeline_schedulers = {
     "ipndm": IPNDMScheduler,
     "k-dpm-2-a": KDPM2AncestralDiscreteScheduler,
     "k-dpm-2": KDPM2DiscreteScheduler,
-    "karras-ve": KarrasVeScheduler,
     "lcm": LCMScheduler,
     "lms-discrete": LMSDiscreteScheduler,
     "pndm": PNDMScheduler,
diff --git a/api/onnx_web/diffusers/version_safe_diffusers.py b/api/onnx_web/diffusers/version_safe_diffusers.py
index d256d615..8c4da406 100644
--- a/api/onnx_web/diffusers/version_safe_diffusers.py
+++ b/api/onnx_web/diffusers/version_safe_diffusers.py
@@ -12,6 +12,11 @@ try:
 except ImportError:
     from ..diffusers.stub_scheduler import StubScheduler as DEISMultistepScheduler
 
+try:
+    from diffusers import DPMSolverSDEScheduler
+except ImportError:
+    from ..diffusers.stub_scheduler import StubScheduler as DPMSolverSDEScheduler
+
 try:
     from diffusers import LCMScheduler
 except ImportError:
diff --git a/api/onnx_web/server/context.py b/api/onnx_web/server/context.py
index 034fc3c6..e02cebd5 100644
--- a/api/onnx_web/server/context.py
+++ b/api/onnx_web/server/context.py
@@ -94,45 +94,44 @@ class ServerContext:
         self.cache = ModelCache(self.cache_limit)
 
     @classmethod
-    def from_environ(cls):
-        memory_limit = environ.get("ONNX_WEB_MEMORY_LIMIT", None)
+    def from_environ(cls, env=environ):
+        memory_limit = env.get("ONNX_WEB_MEMORY_LIMIT", None)
         if memory_limit is not None:
             memory_limit = int(memory_limit)
 
         return cls(
-            bundle_path=environ.get(
-                "ONNX_WEB_BUNDLE_PATH", path.join("..", "gui", "out")
-            ),
-            model_path=environ.get("ONNX_WEB_MODEL_PATH", path.join("..", "models")),
-            output_path=environ.get("ONNX_WEB_OUTPUT_PATH", path.join("..", "outputs")),
-            params_path=environ.get("ONNX_WEB_PARAMS_PATH", "."),
-            cors_origin=get_list(environ, "ONNX_WEB_CORS_ORIGIN", default="*"),
+            bundle_path=env.get("ONNX_WEB_BUNDLE_PATH", path.join("..", "gui", "out")),
+            model_path=env.get("ONNX_WEB_MODEL_PATH", path.join("..", "models")),
+            output_path=env.get("ONNX_WEB_OUTPUT_PATH", path.join("..", "outputs")),
+            params_path=env.get("ONNX_WEB_PARAMS_PATH", "."),
+            cors_origin=get_list(env, "ONNX_WEB_CORS_ORIGIN", default="*"),
             any_platform=get_boolean(
-                environ, "ONNX_WEB_ANY_PLATFORM", DEFAULT_ANY_PLATFORM
+                env, "ONNX_WEB_ANY_PLATFORM", DEFAULT_ANY_PLATFORM
             ),
-            block_platforms=get_list(environ, "ONNX_WEB_BLOCK_PLATFORMS"),
-            default_platform=environ.get("ONNX_WEB_DEFAULT_PLATFORM", None),
-            image_format=environ.get("ONNX_WEB_IMAGE_FORMAT", DEFAULT_IMAGE_FORMAT),
-            cache_limit=int(environ.get("ONNX_WEB_CACHE_MODELS", DEFAULT_CACHE_LIMIT)),
+            block_platforms=get_list(env, "ONNX_WEB_BLOCK_PLATFORMS"),
+            default_platform=env.get("ONNX_WEB_DEFAULT_PLATFORM", None),
+            image_format=env.get("ONNX_WEB_IMAGE_FORMAT", DEFAULT_IMAGE_FORMAT),
+            cache_limit=int(env.get("ONNX_WEB_CACHE_MODELS", DEFAULT_CACHE_LIMIT)),
             show_progress=get_boolean(
-                environ, "ONNX_WEB_SHOW_PROGRESS", DEFAULT_SHOW_PROGRESS
+                env, "ONNX_WEB_SHOW_PROGRESS", DEFAULT_SHOW_PROGRESS
             ),
-            optimizations=get_list(environ, "ONNX_WEB_OPTIMIZATIONS"),
-            extra_models=get_list(environ, "ONNX_WEB_EXTRA_MODELS"),
-            job_limit=int(environ.get("ONNX_WEB_JOB_LIMIT", DEFAULT_JOB_LIMIT)),
+            optimizations=get_list(env, "ONNX_WEB_OPTIMIZATIONS"),
+            extra_models=get_list(env, "ONNX_WEB_EXTRA_MODELS"),
+            job_limit=int(env.get("ONNX_WEB_JOB_LIMIT", DEFAULT_JOB_LIMIT)),
             memory_limit=memory_limit,
-            admin_token=environ.get("ONNX_WEB_ADMIN_TOKEN", None),
-            server_version=environ.get(
-                "ONNX_WEB_SERVER_VERSION", DEFAULT_SERVER_VERSION
-            ),
+            admin_token=env.get("ONNX_WEB_ADMIN_TOKEN", None),
+            server_version=env.get("ONNX_WEB_SERVER_VERSION", DEFAULT_SERVER_VERSION),
             worker_retries=int(
-                environ.get("ONNX_WEB_WORKER_RETRIES", DEFAULT_WORKER_RETRIES)
+                env.get("ONNX_WEB_WORKER_RETRIES", DEFAULT_WORKER_RETRIES)
             ),
-            feature_flags=get_list(environ, "ONNX_WEB_FEATURE_FLAGS"),
-            plugins=get_list(environ, "ONNX_WEB_PLUGINS", ""),
-            debug=get_boolean(environ, "ONNX_WEB_DEBUG", False),
+            feature_flags=get_list(env, "ONNX_WEB_FEATURE_FLAGS"),
+            plugins=get_list(env, "ONNX_WEB_PLUGINS", ""),
+            debug=get_boolean(env, "ONNX_WEB_DEBUG", False),
         )
 
+    def get_setting(self, flag: str, default: str) -> Optional[str]:
+        return environ.get(f"ONNX_WEB_{flag}", default)
+
     def has_feature(self, flag: str) -> bool:
         return flag in self.feature_flags
 
diff --git a/api/onnx_web/server/load.py b/api/onnx_web/server/load.py
index 6bf1de2d..2dd37157 100644
--- a/api/onnx_web/server/load.py
+++ b/api/onnx_web/server/load.py
@@ -8,6 +8,7 @@ from typing import Any, Dict, List, Optional, Union
 import torch
 from jsonschema import ValidationError, validate
 
+from ..convert.utils import fix_diffusion_name
 from ..image import (  # mask filters; noise sources
     mask_filter_gaussian_multiply,
     mask_filter_gaussian_screen,
@@ -189,6 +190,9 @@ def load_extras(server: ServerContext):
                         for model in data[model_type]:
                             model_name = model["name"]
 
+                            if model_type == "diffusion":
+                                model_name = fix_diffusion_name(model_name)
+
                             if "hash" in model:
                                 logger.debug(
                                     "collecting hash for model %s from %s",
diff --git a/api/onnx_web/server/plugin.py b/api/onnx_web/server/plugin.py
index 022047df..2adcabb3 100644
--- a/api/onnx_web/server/plugin.py
+++ b/api/onnx_web/server/plugin.py
@@ -2,18 +2,24 @@ from importlib import import_module
 from logging import getLogger
 from typing import Any, Callable, Dict
 
-from onnx_web.chain.stages import add_stage
-from onnx_web.diffusers.load import add_pipeline
-from onnx_web.server.context import ServerContext
+from ..chain.stages import add_stage
+from ..diffusers.load import add_pipeline
+from ..server.context import ServerContext
 
 logger = getLogger(__name__)
 
 
 class PluginExports:
+    clients: Dict[str, Any]
+    converter: Dict[str, Any]
     pipelines: Dict[str, Any]
     stages: Dict[str, Any]
 
-    def __init__(self, pipelines=None, stages=None) -> None:
+    def __init__(
+        self, clients=None, converter=None, pipelines=None, stages=None
+    ) -> None:
+        self.clients = clients or {}
+        self.converter = converter or {}
         self.pipelines = pipelines or {}
         self.stages = stages or {}
 
diff --git a/api/params.json b/api/params.json
index d9cd6fa0..7b0c746f 100644
--- a/api/params.json
+++ b/api/params.json
@@ -14,7 +14,7 @@
   },
   "cfg": {
     "default": 6,
-    "min": 1,
+    "min": 0,
     "max": 30,
     "step": 0.1
   },
diff --git a/api/requirements/base.txt b/api/requirements/base.txt
index 9d2b8ddf..822ef4e7 100644
--- a/api/requirements/base.txt
+++ b/api/requirements/base.txt
@@ -3,21 +3,21 @@ numpy==1.23.5
 protobuf==3.20.3
 
 ### SD packages ###
-accelerate==0.22.0
+accelerate==0.25.0
 coloredlogs==15.0.1
 controlnet_aux==0.0.2
-datasets==2.14.3
-diffusers==0.20.0
+datasets==2.15.0
+diffusers==0.24.0
 huggingface-hub==0.16.4
 invisible-watermark==0.2.0
 mediapipe==0.9.2.1
 omegaconf==2.3.0
 onnx==1.13.0
 # onnxruntime has many platform-specific packages
-optimum==1.12.0
-safetensors==0.3.1
-timm==0.6.13
-transformers==4.32.0
+optimum==1.16.0
+safetensors==0.4.1
+timm==0.9.12
+transformers==4.36.1
 
 #### Upscaling and face correction
 basicsr==1.4.2
@@ -29,10 +29,11 @@ realesrgan==0.3.0
 ### Server packages ###
 arpeggio==2.0.0
 boto3==1.26.69
-flask==2.2.2
+flask==3.0.0
 flask-cors==3.0.10
 jsonschema==4.17.3
 piexif==1.1.3
 pyyaml==6.0
 setproctitle==1.3.2
 waitress==2.1.2
+werkzeug==3.0.1
diff --git a/api/tests/convert/test_utils.py b/api/tests/convert/test_utils.py
index 4281adbc..34b0bf9b 100644
--- a/api/tests/convert/test_utils.py
+++ b/api/tests/convert/test_utils.py
@@ -27,7 +27,7 @@ class ConversionContextTests(unittest.TestCase):
 
 class DownloadProgressTests(unittest.TestCase):
     def test_download_example(self):
-        path = download_progress([("https://example.com", "/tmp/example-dot-com")])
+        path = download_progress("https://example.com", "/tmp/example-dot-com")
         self.assertEqual(path, "/tmp/example-dot-com")
 
 
diff --git a/api/tests/test_diffusers/test_run.py b/api/tests/test_diffusers/test_run.py
index 26578f3e..322712e4 100644
--- a/api/tests/test_diffusers/test_run.py
+++ b/api/tests/test_diffusers/test_run.py
@@ -414,6 +414,7 @@ class TestBlendPipeline(unittest.TestCase):
                 3.0,
                 1,
                 1,
+                unet_tile=64,
             ),
             Size(64, 64),
             ["test-blend.png"],
diff --git a/docs/getting-started.md b/docs/getting-started.md
new file mode 100644
index 00000000..fc596089
--- /dev/null
+++ b/docs/getting-started.md
@@ -0,0 +1,198 @@
+# Getting Started With onnx-web
+
+onnx-web is a tool for generating images with Stable Diffusion pipelines, including SDXL.
+
+## Contents
+
+- [Getting Started With onnx-web](#getting-started-with-onnx-web)
+  - [Contents](#contents)
+  - [Setup](#setup)
+    - [Windows bundle setup](#windows-bundle-setup)
+    - [Other setup methods](#other-setup-methods)
+  - [Running](#running)
+    - [Running the server](#running-the-server)
+    - [Running the web UI](#running-the-web-ui)
+  - [Tabs](#tabs)
+    - [Txt2img Tab](#txt2img-tab)
+    - [Img2img Tab](#img2img-tab)
+    - [Inpaint Tab](#inpaint-tab)
+    - [Upscale Tab](#upscale-tab)
+    - [Blend Tab](#blend-tab)
+    - [Models Tab](#models-tab)
+    - [Settings Tab](#settings-tab)
+  - [Image parameters](#image-parameters)
+    - [Common image parameters](#common-image-parameters)
+    - [Unique image parameters](#unique-image-parameters)
+  - [Prompt syntax](#prompt-syntax)
+    - [LoRAs and embeddings](#loras-and-embeddings)
+    - [CLIP skip](#clip-skip)
+  - [Highres](#highres)
+    - [Highres prompt](#highres-prompt)
+    - [Highres iterations](#highres-iterations)
+  - [Profiles](#profiles)
+    - [Loading from files](#loading-from-files)
+    - [Saving profiles in the web UI](#saving-profiles-in-the-web-ui)
+    - [Sharing parameters profiles](#sharing-parameters-profiles)
+  - [Panorama pipeline](#panorama-pipeline)
+    - [Region prompts](#region-prompts)
+    - [Region seeds](#region-seeds)
+  - [Grid mode](#grid-mode)
+    - [Grid tokens](#grid-tokens)
+  - [Memory optimizations](#memory-optimizations)
+    - [Converting to fp16](#converting-to-fp16)
+    - [Moving models to the CPU](#moving-models-to-the-cpu)
+
+## Setup
+
+### Windows bundle setup
+
+1. Download
+2. Extract
+3. Security flags
+4. Run
+
+### Other setup methods
+
+Link to the other methods.
+
+## Running
+
+### Running the server
+
+Run server or bundle.
+
+### Running the web UI
+
+Open web UI.
+
+Use it from your phone.
+
+## Tabs
+
+There are 5 tabs, which do different things.
+
+### Txt2img Tab
+
+Words go in, pictures come out.
+
+### Img2img Tab
+
+Pictures go in, better pictures come out.
+
+ControlNet lives here.
+
+### Inpaint Tab
+
+Pictures go in, parts of the same picture come out.
+
+### Upscale Tab
+
+Just highres and super resolution.
+
+### Blend Tab
+
+Use the mask tool to combine two images.
+
+### Models Tab
+
+Add and manage models.
+
+### Settings Tab
+
+Manage web UI settings.
+
+Reset buttons.
+
+## Image parameters
+
+### Common image parameters
+
+- Scheduler
+- Eta
+  - for DDIM
+- CFG
+- Steps
+- Seed
+- Batch size
+- Prompt
+- Negative prompt
+- Width, height
+
+### Unique image parameters
+
+- UNet tile size
+- UNet overlap
+- Tiled VAE
+- VAE tile size
+- VAE overlap
+
+See the complete user guide for details about the highres, upscale, and correction parameters.
+
+## Prompt syntax
+
+### LoRAs and embeddings
+
+`<lora:filename:1.0>` and `<inversion:filename:1.0>`.
+
+### CLIP skip
+
+`<clip:skip:2>` for anime.
+
+## Highres
+
+### Highres prompt
+
+`txt2img prompt || img2img prompt`
+
+### Highres iterations
+
+Highres will apply the upscaler and highres prompt (img2img pipeline) for each iteration.
+
+The final size will be `scale ** iterations`.
+
+## Profiles
+
+Saved sets of parameters for later use.
+
+### Loading from files
+
+- load from images
+- load from JSON
+
+### Saving profiles in the web UI
+
+Use the save button.
+
+### Sharing parameters profiles
+
+Use the download button.
+
+Share profiles in the Discord channel.
+
+## Panorama pipeline
+
+### Region prompts
+
+`<region:X:Y:W:H:S:F_TLBR:prompt+>`
+
+### Region seeds
+
+`<reseed:X:Y:W:H:?:F_TLBR:seed>`
+
+## Grid mode
+
+Makes many images. Takes many time.
+
+### Grid tokens
+
+`__column__` and `__row__` if you pick token in the menu.
+
+## Memory optimizations
+
+### Converting to fp16
+
+Enable the option.
+
+### Moving models to the CPU
+
+Option for each model.
diff --git a/gui/src/components/ImageHistory.tsx b/gui/src/components/ImageHistory.tsx
index 0f6b0f2f..20a520ed 100644
--- a/gui/src/components/ImageHistory.tsx
+++ b/gui/src/components/ImageHistory.tsx
@@ -6,7 +6,7 @@ import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
-import { OnnxState, StateContext } from '../state.js';
+import { OnnxState, StateContext } from '../state/full.js';
 import { ErrorCard } from './card/ErrorCard.js';
 import { ImageCard } from './card/ImageCard.js';
 import { LoadingCard } from './card/LoadingCard.js';
diff --git a/gui/src/components/OnnxError.tsx b/gui/src/components/OnnxError.tsx
index da6d7147..05186768 100644
--- a/gui/src/components/OnnxError.tsx
+++ b/gui/src/components/OnnxError.tsx
@@ -2,7 +2,7 @@ import { Box, Button, Container, Stack, Typography } from '@mui/material';
 import * as React from 'react';
 import { ReactNode } from 'react';
 
-import { STATE_KEY } from '../state.js';
+import { STATE_KEY } from '../state/full.js';
 import { Logo } from './Logo.js';
 
 export interface OnnxErrorProps {
diff --git a/gui/src/components/OnnxWeb.tsx b/gui/src/components/OnnxWeb.tsx
index 918ee84f..69c2db20 100644
--- a/gui/src/components/OnnxWeb.tsx
+++ b/gui/src/components/OnnxWeb.tsx
@@ -7,7 +7,7 @@ import { useContext, useMemo } from 'react';
 import { useHash } from 'react-use/lib/useHash';
 import { useStore } from 'zustand';
 
-import { OnnxState, StateContext } from '../state.js';
+import { OnnxState, StateContext } from '../state/full.js';
 import { ImageHistory } from './ImageHistory.js';
 import { Logo } from './Logo.js';
 import { Blend } from './tab/Blend.js';
diff --git a/gui/src/components/Profiles.tsx b/gui/src/components/Profiles.tsx
index baf022b3..de8f2263 100644
--- a/gui/src/components/Profiles.tsx
+++ b/gui/src/components/Profiles.tsx
@@ -21,7 +21,7 @@ import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
-import { OnnxState, StateContext } from '../state.js';
+import { OnnxState, StateContext } from '../state/full.js';
 import { ImageMetadata } from '../types/api.js';
 import { DeepPartial } from '../types/model.js';
 import { BaseImgParams, HighresParams, Txt2ImgParams, UpscaleParams } from '../types/params.js';
diff --git a/gui/src/components/card/ErrorCard.tsx b/gui/src/components/card/ErrorCard.tsx
index bb3ac6c9..f7106584 100644
--- a/gui/src/components/card/ErrorCard.tsx
+++ b/gui/src/components/card/ErrorCard.tsx
@@ -9,7 +9,7 @@ import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
-import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state.js';
+import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state/full.js';
 import { ImageResponse, ReadyResponse, RetryParams } from '../../types/api.js';
 
 export interface ErrorCardProps {
diff --git a/gui/src/components/card/ImageCard.tsx b/gui/src/components/card/ImageCard.tsx
index 7c35acb2..e614d430 100644
--- a/gui/src/components/card/ImageCard.tsx
+++ b/gui/src/components/card/ImageCard.tsx
@@ -8,9 +8,10 @@ import { useHash } from 'react-use/lib/useHash';
 import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
-import { BLEND_SOURCES, ConfigContext, OnnxState, StateContext } from '../../state.js';
+import { ConfigContext, OnnxState, StateContext } from '../../state/full.js';
 import { ImageResponse } from '../../types/api.js';
 import { range, visibleIndex } from '../../utils.js';
+import { BLEND_SOURCES } from '../../constants.js';
 
 export interface ImageCardProps {
   image: ImageResponse;
diff --git a/gui/src/components/card/LoadingCard.tsx b/gui/src/components/card/LoadingCard.tsx
index e0fcdb68..71bfb5f0 100644
--- a/gui/src/components/card/LoadingCard.tsx
+++ b/gui/src/components/card/LoadingCard.tsx
@@ -9,7 +9,7 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { POLL_TIME } from '../../config.js';
-import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state.js';
+import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state/full.js';
 import { ImageResponse } from '../../types/api.js';
 
 const LOADING_PERCENT = 100;
diff --git a/gui/src/components/control/HighresControl.tsx b/gui/src/components/control/HighresControl.tsx
index 91525b21..62fa63fd 100644
--- a/gui/src/components/control/HighresControl.tsx
+++ b/gui/src/components/control/HighresControl.tsx
@@ -5,7 +5,7 @@ import { useContext } from 'react';
 import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 
-import { ConfigContext, OnnxState, StateContext } from '../../state.js';
+import { ConfigContext, OnnxState, StateContext } from '../../state/full.js';
 import { HighresParams } from '../../types/params.js';
 import { NumericField } from '../input/NumericField.js';
 
diff --git a/gui/src/components/control/ImageControl.tsx b/gui/src/components/control/ImageControl.tsx
index 8271c700..8877ae50 100644
--- a/gui/src/components/control/ImageControl.tsx
+++ b/gui/src/components/control/ImageControl.tsx
@@ -11,7 +11,7 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { STALE_TIME } from '../../config.js';
-import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state.js';
+import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state/full.js';
 import { BaseImgParams } from '../../types/params.js';
 import { NumericField } from '../input/NumericField.js';
 import { PromptInput } from '../input/PromptInput.js';
diff --git a/gui/src/components/control/ModelControl.tsx b/gui/src/components/control/ModelControl.tsx
index ba08998f..54b99846 100644
--- a/gui/src/components/control/ModelControl.tsx
+++ b/gui/src/components/control/ModelControl.tsx
@@ -6,7 +6,7 @@ import { useContext } from 'react';
 import { useTranslation } from 'react-i18next';
 
 import { STALE_TIME } from '../../config.js';
-import { ClientContext } from '../../state.js';
+import { ClientContext } from '../../state/full.js';
 import { ModelParams } from '../../types/params.js';
 import { QueryList } from '../input/QueryList.js';
 
diff --git a/gui/src/components/control/OutpaintControl.tsx b/gui/src/components/control/OutpaintControl.tsx
index 5b70cc50..69523477 100644
--- a/gui/src/components/control/OutpaintControl.tsx
+++ b/gui/src/components/control/OutpaintControl.tsx
@@ -6,7 +6,7 @@ import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
-import { ConfigContext, OnnxState, StateContext } from '../../state.js';
+import { ConfigContext, OnnxState, StateContext } from '../../state/full.js';
 import { NumericField } from '../input/NumericField.js';
 
 export function OutpaintControl() {
diff --git a/gui/src/components/control/UpscaleControl.tsx b/gui/src/components/control/UpscaleControl.tsx
index a90dd330..5d840a04 100644
--- a/gui/src/components/control/UpscaleControl.tsx
+++ b/gui/src/components/control/UpscaleControl.tsx
@@ -5,7 +5,7 @@ import { useContext } from 'react';
 import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 
-import { ConfigContext, OnnxState, StateContext } from '../../state.js';
+import { ConfigContext, OnnxState, StateContext } from '../../state/full.js';
 import { UpscaleParams } from '../../types/params.js';
 import { NumericField } from '../input/NumericField.js';
 
diff --git a/gui/src/components/control/VariableControl.tsx b/gui/src/components/control/VariableControl.tsx
index 32a66660..bf5139e1 100644
--- a/gui/src/components/control/VariableControl.tsx
+++ b/gui/src/components/control/VariableControl.tsx
@@ -5,7 +5,7 @@ import { useContext } from 'react';
 import { useStore } from 'zustand';
 
 import { PipelineGrid } from '../../client/utils.js';
-import { OnnxState, StateContext } from '../../state.js';
+import { OnnxState, StateContext } from '../../state/full.js';
 import { VARIABLE_PARAMETERS } from '../../types/chain.js';
 
 export interface VariableControlProps {
diff --git a/gui/src/components/input/EditableList.tsx b/gui/src/components/input/EditableList.tsx
index 3910e394..a6d45aca 100644
--- a/gui/src/components/input/EditableList.tsx
+++ b/gui/src/components/input/EditableList.tsx
@@ -4,7 +4,7 @@ import * as React from 'react';
 import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 
-import { OnnxState, StateContext } from '../../state.js';
+import { OnnxState, StateContext } from '../../state/full.js';
 
 const { useContext, useState, memo, useMemo } = React;
 
diff --git a/gui/src/components/input/MaskCanvas.tsx b/gui/src/components/input/MaskCanvas.tsx
index ae7f2723..2e605423 100644
--- a/gui/src/components/input/MaskCanvas.tsx
+++ b/gui/src/components/input/MaskCanvas.tsx
@@ -6,7 +6,7 @@ import React, { RefObject, useContext, useEffect, useMemo, useRef } from 'react'
 import { useTranslation } from 'react-i18next';
 
 import { SAVE_TIME } from '../../config.js';
-import { ConfigContext, LoggerContext, StateContext } from '../../state.js';
+import { ConfigContext, LoggerContext, StateContext } from '../../state/full.js';
 import { BrushParams } from '../../types/params.js';
 import { imageFromBlob } from '../../utils.js';
 import { NumericField } from './NumericField';
diff --git a/gui/src/components/input/PromptInput.tsx b/gui/src/components/input/PromptInput.tsx
index cdaa6751..bc522569 100644
--- a/gui/src/components/input/PromptInput.tsx
+++ b/gui/src/components/input/PromptInput.tsx
@@ -8,7 +8,7 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { STALE_TIME } from '../../config.js';
-import { ClientContext, OnnxState, StateContext } from '../../state.js';
+import { ClientContext, OnnxState, StateContext } from '../../state/full.js';
 import { QueryMenu } from '../input/QueryMenu.js';
 import { ModelResponse } from '../../types/api.js';
 
diff --git a/gui/src/components/tab/Blend.tsx b/gui/src/components/tab/Blend.tsx
index a1a304d1..21fe47b6 100644
--- a/gui/src/components/tab/Blend.tsx
+++ b/gui/src/components/tab/Blend.tsx
@@ -8,7 +8,9 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { IMAGE_FILTER } from '../../config.js';
-import { BLEND_SOURCES, ClientContext, OnnxState, StateContext, TabState } from '../../state.js';
+import { BLEND_SOURCES } from '../../constants.js';
+import { ClientContext, OnnxState, StateContext } from '../../state/full.js';
+import { TabState } from '../../state/types.js';
 import { BlendParams, BrushParams, ModelParams, UpscaleParams } from '../../types/params.js';
 import { range } from '../../utils.js';
 import { UpscaleControl } from '../control/UpscaleControl.js';
diff --git a/gui/src/components/tab/Img2Img.tsx b/gui/src/components/tab/Img2Img.tsx
index cddb8539..26bda5cf 100644
--- a/gui/src/components/tab/Img2Img.tsx
+++ b/gui/src/components/tab/Img2Img.tsx
@@ -8,7 +8,8 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { IMAGE_FILTER, STALE_TIME } from '../../config.js';
-import { ClientContext, ConfigContext, OnnxState, StateContext, TabState } from '../../state.js';
+import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state/full.js';
+import { TabState } from '../../state/types.js';
 import { HighresParams, Img2ImgParams, ModelParams, UpscaleParams } from '../../types/params.js';
 import { Profiles } from '../Profiles.js';
 import { HighresControl } from '../control/HighresControl.js';
diff --git a/gui/src/components/tab/Inpaint.tsx b/gui/src/components/tab/Inpaint.tsx
index b783f5b5..ca83835b 100644
--- a/gui/src/components/tab/Inpaint.tsx
+++ b/gui/src/components/tab/Inpaint.tsx
@@ -8,7 +8,8 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { IMAGE_FILTER, STALE_TIME } from '../../config.js';
-import { ClientContext, ConfigContext, OnnxState, StateContext, TabState } from '../../state.js';
+import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state/full.js';
+import { TabState } from '../../state/types.js';
 import { BrushParams, HighresParams, InpaintParams, ModelParams, UpscaleParams } from '../../types/params.js';
 import { Profiles } from '../Profiles.js';
 import { HighresControl } from '../control/HighresControl.js';
diff --git a/gui/src/components/tab/Models.tsx b/gui/src/components/tab/Models.tsx
index e634485a..86e96ecd 100644
--- a/gui/src/components/tab/Models.tsx
+++ b/gui/src/components/tab/Models.tsx
@@ -8,7 +8,7 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { STALE_TIME } from '../../config.js';
-import { ClientContext, OnnxState, StateContext } from '../../state.js';
+import { ClientContext, OnnxState, StateContext } from '../../state/full.js';
 import {
   CorrectionModel,
   DiffusionModel,
diff --git a/gui/src/components/tab/Settings.tsx b/gui/src/components/tab/Settings.tsx
index 5f09c1a2..7b25ed3b 100644
--- a/gui/src/components/tab/Settings.tsx
+++ b/gui/src/components/tab/Settings.tsx
@@ -7,7 +7,7 @@ import { useTranslation } from 'react-i18next';
 import { useStore } from 'zustand';
 
 import { getApiRoot } from '../../config.js';
-import { ConfigContext, StateContext, STATE_KEY } from '../../state.js';
+import { ConfigContext, StateContext, STATE_KEY } from '../../state/full.js';
 import { getTheme } from '../utils.js';
 import { NumericField } from '../input/NumericField.js';
 
diff --git a/gui/src/components/tab/Txt2Img.tsx b/gui/src/components/tab/Txt2Img.tsx
index 921def1f..81286ec2 100644
--- a/gui/src/components/tab/Txt2Img.tsx
+++ b/gui/src/components/tab/Txt2Img.tsx
@@ -8,7 +8,8 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { PipelineGrid, makeTxt2ImgGridPipeline } from '../../client/utils.js';
-import { ClientContext, ConfigContext, OnnxState, StateContext, TabState } from '../../state.js';
+import { ClientContext, ConfigContext, OnnxState, StateContext } from '../../state/full.js';
+import { TabState } from '../../state/types.js';
 import { HighresParams, ModelParams, Txt2ImgParams, UpscaleParams } from '../../types/params.js';
 import { Profiles } from '../Profiles.js';
 import { HighresControl } from '../control/HighresControl.js';
diff --git a/gui/src/components/tab/Upscale.tsx b/gui/src/components/tab/Upscale.tsx
index 06314579..e9a1c482 100644
--- a/gui/src/components/tab/Upscale.tsx
+++ b/gui/src/components/tab/Upscale.tsx
@@ -8,7 +8,8 @@ import { useStore } from 'zustand';
 import { shallow } from 'zustand/shallow';
 
 import { IMAGE_FILTER } from '../../config.js';
-import { ClientContext, OnnxState, StateContext, TabState } from '../../state.js';
+import { ClientContext, OnnxState, StateContext } from '../../state/full.js';
+import { TabState } from '../../state/types.js';
 import { HighresParams, ModelParams, UpscaleParams, UpscaleReqParams } from '../../types/params.js';
 import { Profiles } from '../Profiles.js';
 import { HighresControl } from '../control/HighresControl.js';
diff --git a/gui/src/components/utils.ts b/gui/src/components/utils.ts
index 58e106ec..39ba9bae 100644
--- a/gui/src/components/utils.ts
+++ b/gui/src/components/utils.ts
@@ -1,6 +1,6 @@
 import { PaletteMode } from '@mui/material';
 
-import { Theme } from '../state.js';
+import { Theme } from '../state/types.js';
 import { trimHash } from '../utils.js';
 
 export const TAB_LABELS = [
diff --git a/gui/src/config.json b/gui/src/config.json
index 91cccb8e..d2d63098 100644
--- a/gui/src/config.json
+++ b/gui/src/config.json
@@ -18,7 +18,7 @@
     },
     "cfg": {
       "default": 6,
-      "min": 1,
+      "min": 0,
       "max": 30,
       "step": 0.1
     },
diff --git a/gui/src/constants.ts b/gui/src/constants.ts
new file mode 100644
index 00000000..eb718447
--- /dev/null
+++ b/gui/src/constants.ts
@@ -0,0 +1,31 @@
+
+export const BLEND_SOURCES = 2;
+
+/**
+ * Default parameters for the inpaint brush.
+ *
+ * Not provided by the server yet.
+ */
+export const DEFAULT_BRUSH = {
+  color: 255,
+  size: 8,
+  strength: 0.5,
+};
+
+/**
+ * Default parameters for the image history.
+ *
+ * Not provided by the server yet.
+ */
+export const DEFAULT_HISTORY = {
+  /**
+   * The number of images to be shown.
+   */
+  limit: 4,
+
+  /**
+   * The number of additional images to be kept in history, so they can scroll
+   * back into view when you delete one. Does not include deleted images.
+   */
+  scrollback: 2,
+};
diff --git a/gui/src/main.tsx b/gui/src/main.tsx
index d0b83a11..81aeecf0 100644
--- a/gui/src/main.tsx
+++ b/gui/src/main.tsx
@@ -28,8 +28,9 @@ import {
   STATE_KEY,
   STATE_VERSION,
   StateContext,
-} from './state.js';
+} from './state/full.js';
 import { I18N_STRINGS } from './strings/all.js';
+import { applyStateMigrations, UnknownState } from './state/migration/default.js';
 
 export const INITIAL_LOAD_TIMEOUT = 5_000;
 
@@ -70,6 +71,9 @@ export async function renderApp(config: Config, params: ServerParams, logger: Lo
     ...createResetSlice(...slice),
     ...createProfileSlice(...slice),
   }), {
+    migrate(persistedState, version) {
+      return applyStateMigrations(params, persistedState as UnknownState, version, logger);
+    },
     name: STATE_KEY,
     partialize(s) {
       return {
diff --git a/gui/src/state.ts b/gui/src/state.ts
deleted file mode 100644
index ad2a3e3a..00000000
--- a/gui/src/state.ts
+++ /dev/null
@@ -1,965 +0,0 @@
-/* eslint-disable camelcase */
-/* eslint-disable max-lines */
-/* eslint-disable no-null/no-null */
-import { Maybe } from '@apextoaster/js-utils';
-import { PaletteMode } from '@mui/material';
-import { Logger } from 'noicejs';
-import { createContext } from 'react';
-import { StateCreator, StoreApi } from 'zustand';
-
-import {
-  ApiClient,
-} from './client/base.js';
-import { PipelineGrid } from './client/utils.js';
-import { Config, ConfigFiles, ConfigState, ServerParams } from './config.js';
-import { CorrectionModel, DiffusionModel, ExtraNetwork, ExtraSource, ExtrasFile, UpscalingModel } from './types/model.js';
-import { ImageResponse, ReadyResponse, RetryParams } from './types/api.js';
-import {
-  BaseImgParams,
-  BlendParams,
-  BrushParams,
-  HighresParams,
-  Img2ImgParams,
-  InpaintParams,
-  ModelParams,
-  OutpaintPixels,
-  Txt2ImgParams,
-  UpscaleParams,
-  UpscaleReqParams,
-} from './types/params.js';
-
-export const MISSING_INDEX = -1;
-
-export type Theme = PaletteMode | ''; // tri-state, '' is unset
-
-/**
- * Combine optional files and required ranges.
- */
-export type TabState<TabParams> = ConfigFiles<Required<TabParams>> & ConfigState<Required<TabParams>>;
-
-export interface HistoryItem {
-  image: ImageResponse;
-  ready: Maybe<ReadyResponse>;
-  retry: Maybe<RetryParams>;
-}
-
-export interface ProfileItem {
-  name: string;
-  params: BaseImgParams | Txt2ImgParams;
-  highres?: Maybe<HighresParams>;
-  upscale?: Maybe<UpscaleParams>;
-}
-
-interface DefaultSlice {
-  defaults: TabState<BaseImgParams>;
-  theme: Theme;
-
-  setDefaults(param: Partial<BaseImgParams>): void;
-  setTheme(theme: Theme): void;
-}
-
-interface HistorySlice {
-  history: Array<HistoryItem>;
-  limit: number;
-
-  pushHistory(image: ImageResponse, retry?: RetryParams): void;
-  removeHistory(image: ImageResponse): void;
-  setLimit(limit: number): void;
-  setReady(image: ImageResponse, ready: ReadyResponse): void;
-}
-
-interface ModelSlice {
-  extras: ExtrasFile;
-
-  removeCorrectionModel(model: CorrectionModel): void;
-  removeDiffusionModel(model: DiffusionModel): void;
-  removeExtraNetwork(model: ExtraNetwork): void;
-  removeExtraSource(model: ExtraSource): void;
-  removeUpscalingModel(model: UpscalingModel): void;
-
-  setExtras(extras: Partial<ExtrasFile>): void;
-
-  setCorrectionModel(model: CorrectionModel): void;
-  setDiffusionModel(model: DiffusionModel): void;
-  setExtraNetwork(model: ExtraNetwork): void;
-  setExtraSource(model: ExtraSource): void;
-  setUpscalingModel(model: UpscalingModel): void;
-}
-
-// #region tab slices
-interface Txt2ImgSlice {
-  txt2img: TabState<Txt2ImgParams>;
-  txt2imgModel: ModelParams;
-  txt2imgHighres: HighresParams;
-  txt2imgUpscale: UpscaleParams;
-  txt2imgVariable: PipelineGrid;
-
-  resetTxt2Img(): void;
-
-  setTxt2Img(params: Partial<Txt2ImgParams>): void;
-  setTxt2ImgModel(params: Partial<ModelParams>): void;
-  setTxt2ImgHighres(params: Partial<HighresParams>): void;
-  setTxt2ImgUpscale(params: Partial<UpscaleParams>): void;
-  setTxt2ImgVariable(params: Partial<PipelineGrid>): void;
-}
-
-interface Img2ImgSlice {
-  img2img: TabState<Img2ImgParams>;
-  img2imgModel: ModelParams;
-  img2imgHighres: HighresParams;
-  img2imgUpscale: UpscaleParams;
-
-  resetImg2Img(): void;
-
-  setImg2Img(params: Partial<Img2ImgParams>): void;
-  setImg2ImgModel(params: Partial<ModelParams>): void;
-  setImg2ImgHighres(params: Partial<HighresParams>): void;
-  setImg2ImgUpscale(params: Partial<UpscaleParams>): void;
-}
-
-interface InpaintSlice {
-  inpaint: TabState<InpaintParams>;
-  inpaintBrush: BrushParams;
-  inpaintModel: ModelParams;
-  inpaintHighres: HighresParams;
-  inpaintUpscale: UpscaleParams;
-  outpaint: OutpaintPixels;
-
-  resetInpaint(): void;
-
-  setInpaint(params: Partial<InpaintParams>): void;
-  setInpaintBrush(brush: Partial<BrushParams>): void;
-  setInpaintModel(params: Partial<ModelParams>): void;
-  setInpaintHighres(params: Partial<HighresParams>): void;
-  setInpaintUpscale(params: Partial<UpscaleParams>): void;
-  setOutpaint(pixels: Partial<OutpaintPixels>): void;
-}
-
-interface UpscaleSlice {
-  upscale: TabState<UpscaleReqParams>;
-  upscaleHighres: HighresParams;
-  upscaleModel: ModelParams;
-  upscaleUpscale: UpscaleParams;
-
-  resetUpscale(): void;
-
-  setUpscale(params: Partial<UpscaleReqParams>): void;
-  setUpscaleHighres(params: Partial<HighresParams>): void;
-  setUpscaleModel(params: Partial<ModelParams>): void;
-  setUpscaleUpscale(params: Partial<UpscaleParams>): void;
-}
-
-interface BlendSlice {
-  blend: TabState<BlendParams>;
-  blendBrush: BrushParams;
-  blendModel: ModelParams;
-  blendUpscale: UpscaleParams;
-
-  resetBlend(): void;
-
-  setBlend(blend: Partial<BlendParams>): void;
-  setBlendBrush(brush: Partial<BrushParams>): void;
-  setBlendModel(model: Partial<ModelParams>): void;
-  setBlendUpscale(params: Partial<UpscaleParams>): void;
-}
-
-interface ResetSlice {
-  resetAll(): void;
-}
-
-interface ProfileSlice {
-  profiles: Array<ProfileItem>;
-
-  removeProfile(profileName: string): void;
-
-  saveProfile(profile: ProfileItem): void;
-}
-// #endregion
-
-/**
- * Full merged state including all slices.
- */
-export type OnnxState
-  = DefaultSlice
-  & HistorySlice
-  & Img2ImgSlice
-  & InpaintSlice
-  & ModelSlice
-  & Txt2ImgSlice
-  & UpscaleSlice
-  & BlendSlice
-  & ResetSlice
-  & ModelSlice
-  & ProfileSlice;
-
-/**
- * Shorthand for state creator to reduce repeated arguments.
- */
-export type Slice<T> = StateCreator<OnnxState, [], [], T>;
-
-/**
- * React context binding for API client.
- */
-export const ClientContext = createContext<Maybe<ApiClient>>(undefined);
-
-/**
- * React context binding for merged config, including server parameters.
- */
-export const ConfigContext = createContext<Maybe<Config<ServerParams>>>(undefined);
-
-/**
- * React context binding for bunyan logger.
- */
-export const LoggerContext = createContext<Maybe<Logger>>(undefined);
-
-/**
- * React context binding for zustand state store.
- */
-export const StateContext = createContext<Maybe<StoreApi<OnnxState>>>(undefined);
-
-/**
- * Key for zustand persistence, typically local storage.
- */
-export const STATE_KEY = 'onnx-web';
-
-/**
- * Current state version for zustand persistence.
- */
-export const STATE_VERSION = 7;
-
-export const BLEND_SOURCES = 2;
-
-/**
- * Default parameters for the inpaint brush.
- *
- * Not provided by the server yet.
- */
-export const DEFAULT_BRUSH = {
-  color: 255,
-  size: 8,
-  strength: 0.5,
-};
-
-/**
- * Default parameters for the image history.
- *
- * Not provided by the server yet.
- */
-export const DEFAULT_HISTORY = {
-  /**
-   * The number of images to be shown.
-   */
-  limit: 4,
-
-  /**
-   * The number of additional images to be kept in history, so they can scroll
-   * back into view when you delete one. Does not include deleted images.
-   */
-  scrollback: 2,
-};
-
-export function baseParamsFromServer(defaults: ServerParams): Required<BaseImgParams> {
-  return {
-    batch: defaults.batch.default,
-    cfg: defaults.cfg.default,
-    eta: defaults.eta.default,
-    negativePrompt: defaults.negativePrompt.default,
-    prompt: defaults.prompt.default,
-    scheduler: defaults.scheduler.default,
-    steps: defaults.steps.default,
-    seed: defaults.seed.default,
-    tiled_vae: defaults.tiled_vae.default,
-    unet_overlap: defaults.unet_overlap.default,
-    unet_tile: defaults.unet_tile.default,
-    vae_overlap: defaults.vae_overlap.default,
-    vae_tile: defaults.vae_tile.default,
-  };
-}
-
-/**
- * Prepare the state slice constructors.
- *
- * In the default state, image sources should be null and booleans should be false. Everything
- * else should be initialized from the default value in the base parameters.
- */
-export function createStateSlices(server: ServerParams) {
-  const defaultParams = baseParamsFromServer(server);
-  const defaultHighres: HighresParams = {
-    enabled: false,
-    highresIterations: server.highresIterations.default,
-    highresMethod: '',
-    highresSteps: server.highresSteps.default,
-    highresScale: server.highresScale.default,
-    highresStrength: server.highresStrength.default,
-  };
-  const defaultModel: ModelParams = {
-    control: server.control.default,
-    correction: server.correction.default,
-    model: server.model.default,
-    pipeline: server.pipeline.default,
-    platform: server.platform.default,
-    upscaling: server.upscaling.default,
-  };
-  const defaultUpscale: UpscaleParams = {
-    denoise: server.denoise.default,
-    enabled: false,
-    faces: false,
-    faceOutscale: server.faceOutscale.default,
-    faceStrength: server.faceStrength.default,
-    outscale: server.outscale.default,
-    scale: server.scale.default,
-    upscaleOrder: server.upscaleOrder.default,
-  };
-  const defaultGrid: PipelineGrid = {
-    enabled: false,
-    columns: {
-      parameter: 'seed',
-      value: '',
-    },
-    rows: {
-      parameter: 'seed',
-      value: '',
-    },
-  };
-
-  const createTxt2ImgSlice: Slice<Txt2ImgSlice> = (set) => ({
-    txt2img: {
-      ...defaultParams,
-      width: server.width.default,
-      height: server.height.default,
-    },
-    txt2imgHighres: {
-      ...defaultHighres,
-    },
-    txt2imgModel: {
-      ...defaultModel,
-    },
-    txt2imgUpscale: {
-      ...defaultUpscale,
-    },
-    txt2imgVariable: {
-      ...defaultGrid,
-    },
-    setTxt2Img(params) {
-      set((prev) => ({
-        txt2img: {
-          ...prev.txt2img,
-          ...params,
-        },
-      }));
-    },
-    setTxt2ImgHighres(params) {
-      set((prev) => ({
-        txt2imgHighres: {
-          ...prev.txt2imgHighres,
-          ...params,
-        },
-      }));
-    },
-    setTxt2ImgModel(params) {
-      set((prev) => ({
-        txt2imgModel: {
-          ...prev.txt2imgModel,
-          ...params,
-        },
-      }));
-    },
-    setTxt2ImgUpscale(params) {
-      set((prev) => ({
-        txt2imgUpscale: {
-          ...prev.txt2imgUpscale,
-          ...params,
-        },
-      }));
-    },
-    setTxt2ImgVariable(params) {
-      set((prev) => ({
-        txt2imgVariable: {
-          ...prev.txt2imgVariable,
-          ...params,
-        },
-      }));
-    },
-    resetTxt2Img() {
-      set({
-        txt2img: {
-          ...defaultParams,
-          width: server.width.default,
-          height: server.height.default,
-        },
-      });
-    },
-  });
-
-  const createImg2ImgSlice: Slice<Img2ImgSlice> = (set) => ({
-    img2img: {
-      ...defaultParams,
-      loopback: server.loopback.default,
-      source: null,
-      sourceFilter: '',
-      strength: server.strength.default,
-    },
-    img2imgHighres: {
-      ...defaultHighres,
-    },
-    img2imgModel: {
-      ...defaultModel,
-    },
-    img2imgUpscale: {
-      ...defaultUpscale,
-    },
-    resetImg2Img() {
-      set({
-        img2img: {
-          ...defaultParams,
-          loopback: server.loopback.default,
-          source: null,
-          sourceFilter: '',
-          strength: server.strength.default,
-        },
-      });
-    },
-    setImg2Img(params) {
-      set((prev) => ({
-        img2img: {
-          ...prev.img2img,
-          ...params,
-        },
-      }));
-    },
-    setImg2ImgHighres(params) {
-      set((prev) => ({
-        img2imgHighres: {
-          ...prev.img2imgHighres,
-          ...params,
-        },
-      }));
-    },
-    setImg2ImgModel(params) {
-      set((prev) => ({
-        img2imgModel: {
-          ...prev.img2imgModel,
-          ...params,
-        },
-      }));
-    },
-    setImg2ImgUpscale(params) {
-      set((prev) => ({
-        img2imgUpscale: {
-          ...prev.img2imgUpscale,
-          ...params,
-        },
-      }));
-    },
-  });
-
-  const createInpaintSlice: Slice<InpaintSlice> = (set) => ({
-    inpaint: {
-      ...defaultParams,
-      fillColor: server.fillColor.default,
-      filter: server.filter.default,
-      mask: null,
-      noise: server.noise.default,
-      source: null,
-      strength: server.strength.default,
-      tileOrder: server.tileOrder.default,
-    },
-    inpaintBrush: {
-      ...DEFAULT_BRUSH,
-    },
-    inpaintHighres: {
-      ...defaultHighres,
-    },
-    inpaintModel: {
-      ...defaultModel,
-    },
-    inpaintUpscale: {
-      ...defaultUpscale,
-    },
-    outpaint: {
-      enabled: false,
-      left: server.left.default,
-      right: server.right.default,
-      top: server.top.default,
-      bottom: server.bottom.default,
-    },
-    resetInpaint() {
-      set({
-        inpaint: {
-          ...defaultParams,
-          fillColor: server.fillColor.default,
-          filter: server.filter.default,
-          mask: null,
-          noise: server.noise.default,
-          source: null,
-          strength: server.strength.default,
-          tileOrder: server.tileOrder.default,
-        },
-      });
-    },
-    setInpaint(params) {
-      set((prev) => ({
-        inpaint: {
-          ...prev.inpaint,
-          ...params,
-        },
-      }));
-    },
-    setInpaintBrush(brush) {
-      set((prev) => ({
-        inpaintBrush: {
-          ...prev.inpaintBrush,
-          ...brush,
-        },
-      }));
-    },
-    setInpaintHighres(params) {
-      set((prev) => ({
-        inpaintHighres: {
-          ...prev.inpaintHighres,
-          ...params,
-        },
-      }));
-    },
-    setInpaintModel(params) {
-      set((prev) => ({
-        inpaintModel: {
-          ...prev.inpaintModel,
-          ...params,
-        },
-      }));
-    },
-    setInpaintUpscale(params) {
-      set((prev) => ({
-        inpaintUpscale: {
-          ...prev.inpaintUpscale,
-          ...params,
-        },
-      }));
-    },
-    setOutpaint(pixels) {
-      set((prev) => ({
-        outpaint: {
-          ...prev.outpaint,
-          ...pixels,
-        }
-      }));
-    },
-  });
-
-  const createHistorySlice: Slice<HistorySlice> = (set) => ({
-    history: [],
-    limit: DEFAULT_HISTORY.limit,
-    pushHistory(image, retry) {
-      set((prev) => ({
-        ...prev,
-        history: [
-          {
-            image,
-            ready: undefined,
-            retry,
-          },
-          ...prev.history,
-        ].slice(0, prev.limit + DEFAULT_HISTORY.scrollback),
-      }));
-    },
-    removeHistory(image) {
-      set((prev) => ({
-        ...prev,
-        history: prev.history.filter((it) => it.image.outputs[0].key !== image.outputs[0].key),
-      }));
-    },
-    setLimit(limit) {
-      set((prev) => ({
-        ...prev,
-        limit,
-      }));
-    },
-    setReady(image, ready) {
-      set((prev) => {
-        const history = [...prev.history];
-        const idx = history.findIndex((it) => it.image.outputs[0].key === image.outputs[0].key);
-        if (idx >= 0) {
-          history[idx].ready = ready;
-        } else {
-          // TODO: error
-        }
-
-        return {
-          ...prev,
-          history,
-        };
-      });
-    },
-  });
-
-  const createUpscaleSlice: Slice<UpscaleSlice> = (set) => ({
-    upscale: {
-      ...defaultParams,
-      source: null,
-    },
-    upscaleHighres: {
-      ...defaultHighres,
-    },
-    upscaleModel: {
-      ...defaultModel,
-    },
-    upscaleUpscale: {
-      ...defaultUpscale,
-    },
-    resetUpscale() {
-      set({
-        upscale: {
-          ...defaultParams,
-          source: null,
-        },
-      });
-    },
-    setUpscale(source) {
-      set((prev) => ({
-        upscale: {
-          ...prev.upscale,
-          ...source,
-        },
-      }));
-    },
-    setUpscaleHighres(params) {
-      set((prev) => ({
-        upscaleHighres: {
-          ...prev.upscaleHighres,
-          ...params,
-        },
-      }));
-    },
-    setUpscaleModel(params) {
-      set((prev) => ({
-        upscaleModel: {
-          ...prev.upscaleModel,
-          ...defaultModel,
-        },
-      }));
-    },
-    setUpscaleUpscale(params) {
-      set((prev) => ({
-        upscaleUpscale: {
-          ...prev.upscaleUpscale,
-          ...params,
-        },
-      }));
-    },
-  });
-
-  const createBlendSlice: Slice<BlendSlice> = (set) => ({
-    blend: {
-      mask: null,
-      sources: [],
-    },
-    blendBrush: {
-      ...DEFAULT_BRUSH,
-    },
-    blendModel: {
-      ...defaultModel,
-    },
-    blendUpscale: {
-      ...defaultUpscale,
-    },
-    resetBlend() {
-      set({
-        blend: {
-          mask: null,
-          sources: [],
-        },
-      });
-    },
-    setBlend(blend) {
-      set((prev) => ({
-        blend: {
-          ...prev.blend,
-          ...blend,
-        },
-      }));
-    },
-    setBlendBrush(brush) {
-      set((prev) => ({
-        blendBrush: {
-          ...prev.blendBrush,
-          ...brush,
-        },
-      }));
-    },
-    setBlendModel(model) {
-      set((prev) => ({
-        blendModel: {
-          ...prev.blendModel,
-          ...model,
-        },
-      }));
-    },
-    setBlendUpscale(params) {
-      set((prev) => ({
-        blendUpscale: {
-          ...prev.blendUpscale,
-          ...params,
-        },
-      }));
-    },
-  });
-
-  const createDefaultSlice: Slice<DefaultSlice> = (set) => ({
-    defaults: {
-      ...defaultParams,
-    },
-    theme: '',
-    setDefaults(params) {
-      set((prev) => ({
-        defaults: {
-          ...prev.defaults,
-          ...params,
-        }
-      }));
-    },
-    setTheme(theme) {
-      set((prev) => ({
-        theme,
-      }));
-    }
-  });
-
-  const createResetSlice: Slice<ResetSlice> = (set) => ({
-    resetAll() {
-      set((prev) => {
-        const next = { ...prev };
-        next.resetImg2Img();
-        next.resetInpaint();
-        next.resetTxt2Img();
-        next.resetUpscale();
-        next.resetBlend();
-        return next;
-      });
-    },
-  });
-
-  const createProfileSlice: Slice<ProfileSlice> = (set) => ({
-    profiles: [],
-    saveProfile(profile: ProfileItem) {
-      set((prev) => {
-        const profiles = [...prev.profiles];
-        const idx = profiles.findIndex((it) => it.name === profile.name);
-        if (idx >= 0) {
-          profiles[idx] = profile;
-        } else {
-          profiles.push(profile);
-        }
-        return {
-          ...prev,
-          profiles,
-        };
-      });
-    },
-    removeProfile(profileName: string) {
-      set((prev) => {
-        const profiles = [...prev.profiles];
-        const idx = profiles.findIndex((it) => it.name === profileName);
-        if (idx >= 0) {
-          profiles.splice(idx, 1);
-        }
-        return {
-          ...prev,
-          profiles,
-        };
-      });
-    }
-  });
-
-  // eslint-disable-next-line sonarjs/cognitive-complexity
-  const createModelSlice: Slice<ModelSlice> = (set) => ({
-    extras: {
-      correction: [],
-      diffusion: [],
-      networks: [],
-      sources: [],
-      upscaling: [],
-    },
-    setExtras(extras) {
-      set((prev) => ({
-        ...prev,
-        extras: {
-          ...prev.extras,
-          ...extras,
-        },
-      }));
-    },
-    setCorrectionModel(model) {
-      set((prev) => {
-        const correction = [...prev.extras.correction];
-        const exists = correction.findIndex((it) => model.name === it.name);
-        if (exists === MISSING_INDEX) {
-          correction.push(model);
-        } else {
-          correction[exists] = model;
-        }
-
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            correction,
-          },
-        };
-      });
-    },
-    setDiffusionModel(model) {
-      set((prev) => {
-        const diffusion = [...prev.extras.diffusion];
-        const exists = diffusion.findIndex((it) => model.name === it.name);
-        if (exists === MISSING_INDEX) {
-          diffusion.push(model);
-        } else {
-          diffusion[exists] = model;
-        }
-
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            diffusion,
-          },
-        };
-      });
-    },
-    setExtraNetwork(model) {
-      set((prev) => {
-        const networks = [...prev.extras.networks];
-        const exists = networks.findIndex((it) => model.name === it.name);
-        if (exists === MISSING_INDEX) {
-          networks.push(model);
-        } else {
-          networks[exists] = model;
-        }
-
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            networks,
-          },
-        };
-      });
-    },
-    setExtraSource(model) {
-      set((prev) => {
-        const sources = [...prev.extras.sources];
-        const exists = sources.findIndex((it) => model.name === it.name);
-        if (exists === MISSING_INDEX) {
-          sources.push(model);
-        } else {
-          sources[exists] = model;
-        }
-
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            sources,
-          },
-        };
-      });
-    },
-    setUpscalingModel(model) {
-      set((prev) => {
-        const upscaling = [...prev.extras.upscaling];
-        const exists = upscaling.findIndex((it) => model.name === it.name);
-        if (exists === MISSING_INDEX) {
-          upscaling.push(model);
-        } else {
-          upscaling[exists] = model;
-        }
-
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            upscaling,
-          },
-        };
-      });
-    },
-    removeCorrectionModel(model) {
-      set((prev) => {
-        const correction = prev.extras.correction.filter((it) => model.name !== it.name);;
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            correction,
-          },
-        };
-      });
-
-    },
-    removeDiffusionModel(model) {
-      set((prev) => {
-        const diffusion = prev.extras.diffusion.filter((it) => model.name !== it.name);;
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            diffusion,
-          },
-        };
-      });
-
-    },
-    removeExtraNetwork(model) {
-      set((prev) => {
-        const networks = prev.extras.networks.filter((it) => model.name !== it.name);;
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            networks,
-          },
-        };
-      });
-
-    },
-    removeExtraSource(model) {
-      set((prev) => {
-        const sources = prev.extras.sources.filter((it) => model.name !== it.name);;
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            sources,
-          },
-        };
-      });
-
-    },
-    removeUpscalingModel(model) {
-      set((prev) => {
-        const upscaling = prev.extras.upscaling.filter((it) => model.name !== it.name);;
-        return {
-          ...prev,
-          extras: {
-            ...prev.extras,
-            upscaling,
-          },
-        };
-      });
-    },
-  });
-
-  return {
-    createDefaultSlice,
-    createHistorySlice,
-    createImg2ImgSlice,
-    createInpaintSlice,
-    createTxt2ImgSlice,
-    createUpscaleSlice,
-    createBlendSlice,
-    createResetSlice,
-    createModelSlice,
-    createProfileSlice,
-  };
-}
diff --git a/gui/src/state/blend.ts b/gui/src/state/blend.ts
new file mode 100644
index 00000000..60cd5f88
--- /dev/null
+++ b/gui/src/state/blend.ts
@@ -0,0 +1,85 @@
+import { DEFAULT_BRUSH } from '../constants.js';
+import {
+  BlendParams,
+  BrushParams,
+  ModelParams,
+  UpscaleParams,
+} from '../types/params.js';
+import { Slice, TabState } from './types.js';
+
+export interface BlendSlice {
+  blend: TabState<BlendParams>;
+  blendBrush: BrushParams;
+  blendModel: ModelParams;
+  blendUpscale: UpscaleParams;
+
+  resetBlend(): void;
+
+  setBlend(blend: Partial<BlendParams>): void;
+  setBlendBrush(brush: Partial<BrushParams>): void;
+  setBlendModel(model: Partial<ModelParams>): void;
+  setBlendUpscale(params: Partial<UpscaleParams>): void;
+}
+
+export function createBlendSlice<TState extends BlendSlice>(
+  defaultModel: ModelParams,
+  defaultUpscale: UpscaleParams,
+): Slice<TState, BlendSlice> {
+  return (set) => ({
+    blend: {
+      // eslint-disable-next-line no-null/no-null
+      mask: null,
+      sources: [],
+    },
+    blendBrush: {
+      ...DEFAULT_BRUSH,
+    },
+    blendModel: {
+      ...defaultModel,
+    },
+    blendUpscale: {
+      ...defaultUpscale,
+    },
+    resetBlend() {
+      set((prev) => ({
+        blend: {
+          // eslint-disable-next-line no-null/no-null
+          mask: null,
+          sources: [] as Array<Blob>,
+        },
+      } as Partial<TState>));
+    },
+    setBlend(blend) {
+      set((prev) => ({
+        blend: {
+          ...prev.blend,
+          ...blend,
+        },
+      } as Partial<TState>));
+    },
+    setBlendBrush(brush) {
+      set((prev) => ({
+        blendBrush: {
+          ...prev.blendBrush,
+          ...brush,
+        },
+      } as Partial<TState>));
+    },
+    setBlendModel(model) {
+      set((prev) => ({
+        blendModel: {
+          ...prev.blendModel,
+          ...model,
+        },
+      } as Partial<TState>));
+    },
+    setBlendUpscale(params) {
+      set((prev) => ({
+        blendUpscale: {
+          ...prev.blendUpscale,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+  });
+}
diff --git a/gui/src/state/default.ts b/gui/src/state/default.ts
new file mode 100644
index 00000000..c578263b
--- /dev/null
+++ b/gui/src/state/default.ts
@@ -0,0 +1,34 @@
+import {
+  BaseImgParams,
+} from '../types/params.js';
+import { Slice, TabState, Theme } from './types.js';
+
+export interface DefaultSlice {
+  defaults: TabState<BaseImgParams>;
+  theme: Theme;
+
+  setDefaults(param: Partial<BaseImgParams>): void;
+  setTheme(theme: Theme): void;
+}
+
+export function createDefaultSlice<TState extends DefaultSlice>(defaultParams: Required<BaseImgParams>): Slice<TState, DefaultSlice> {
+  return (set) => ({
+    defaults: {
+      ...defaultParams,
+    },
+    theme: '',
+    setDefaults(params) {
+      set((prev) => ({
+        defaults: {
+          ...prev.defaults,
+          ...params,
+        }
+      } as Partial<TState>));
+    },
+    setTheme(theme) {
+      set((prev) => ({
+        theme,
+      } as Partial<TState>));
+    }
+  });
+}
diff --git a/gui/src/state/full.ts b/gui/src/state/full.ts
new file mode 100644
index 00000000..990161ae
--- /dev/null
+++ b/gui/src/state/full.ts
@@ -0,0 +1,150 @@
+/* eslint-disable camelcase */
+import { Maybe } from '@apextoaster/js-utils';
+import { Logger } from 'noicejs';
+import { createContext } from 'react';
+import { StoreApi } from 'zustand';
+
+import {
+  ApiClient,
+} from '../client/base.js';
+import { PipelineGrid } from '../client/utils.js';
+import { Config, ServerParams } from '../config.js';
+import { BlendSlice, createBlendSlice } from './blend.js';
+import { DefaultSlice, createDefaultSlice } from './default.js';
+import { HistorySlice, createHistorySlice } from './history.js';
+import { Img2ImgSlice, createImg2ImgSlice } from './img2img.js';
+import { InpaintSlice, createInpaintSlice } from './inpaint.js';
+import { ModelSlice, createModelSlice } from './model.js';
+import { ProfileSlice, createProfileSlice } from './profile.js';
+import { ResetSlice, createResetSlice } from './reset.js';
+import { Txt2ImgSlice, createTxt2ImgSlice } from './txt2img.js';
+import { UpscaleSlice, createUpscaleSlice } from './upscale.js';
+import {
+  BaseImgParams,
+  HighresParams,
+  ModelParams,
+  UpscaleParams,
+} from '../types/params.js';
+
+/**
+ * Full merged state including all slices.
+ */
+export type OnnxState
+  = DefaultSlice
+  & HistorySlice
+  & Img2ImgSlice
+  & InpaintSlice
+  & ModelSlice
+  & Txt2ImgSlice
+  & UpscaleSlice
+  & BlendSlice
+  & ResetSlice
+  & ProfileSlice;
+
+/**
+ * React context binding for API client.
+ */
+export const ClientContext = createContext<Maybe<ApiClient>>(undefined);
+
+/**
+ * React context binding for merged config, including server parameters.
+ */
+export const ConfigContext = createContext<Maybe<Config<ServerParams>>>(undefined);
+
+/**
+ * React context binding for bunyan logger.
+ */
+export const LoggerContext = createContext<Maybe<Logger>>(undefined);
+
+/**
+ * React context binding for zustand state store.
+ */
+export const StateContext = createContext<Maybe<StoreApi<OnnxState>>>(undefined);
+
+/**
+ * Key for zustand persistence, typically local storage.
+ */
+export const STATE_KEY = 'onnx-web';
+
+/**
+ * Current state version for zustand persistence.
+ */
+export const STATE_VERSION = 11;
+
+export function baseParamsFromServer(defaults: ServerParams): Required<BaseImgParams> {
+  return {
+    batch: defaults.batch.default,
+    cfg: defaults.cfg.default,
+    eta: defaults.eta.default,
+    negativePrompt: defaults.negativePrompt.default,
+    prompt: defaults.prompt.default,
+    scheduler: defaults.scheduler.default,
+    steps: defaults.steps.default,
+    seed: defaults.seed.default,
+    tiled_vae: defaults.tiled_vae.default,
+    unet_overlap: defaults.unet_overlap.default,
+    unet_tile: defaults.unet_tile.default,
+    vae_overlap: defaults.vae_overlap.default,
+    vae_tile: defaults.vae_tile.default,
+  };
+}
+
+/**
+ * Prepare the state slice constructors.
+ *
+ * In the default state, image sources should be null and booleans should be false. Everything
+ * else should be initialized from the default value in the base parameters.
+ */
+export function createStateSlices(server: ServerParams) {
+  const defaultParams = baseParamsFromServer(server);
+  const defaultHighres: HighresParams = {
+    enabled: false,
+    highresIterations: server.highresIterations.default,
+    highresMethod: '',
+    highresSteps: server.highresSteps.default,
+    highresScale: server.highresScale.default,
+    highresStrength: server.highresStrength.default,
+  };
+  const defaultModel: ModelParams = {
+    control: server.control.default,
+    correction: server.correction.default,
+    model: server.model.default,
+    pipeline: server.pipeline.default,
+    platform: server.platform.default,
+    upscaling: server.upscaling.default,
+  };
+  const defaultUpscale: UpscaleParams = {
+    denoise: server.denoise.default,
+    enabled: false,
+    faces: false,
+    faceOutscale: server.faceOutscale.default,
+    faceStrength: server.faceStrength.default,
+    outscale: server.outscale.default,
+    scale: server.scale.default,
+    upscaleOrder: server.upscaleOrder.default,
+  };
+  const defaultGrid: PipelineGrid = {
+    enabled: false,
+    columns: {
+      parameter: 'seed',
+      value: '',
+    },
+    rows: {
+      parameter: 'seed',
+      value: '',
+    },
+  };
+
+  return {
+    createBlendSlice: createBlendSlice(defaultModel, defaultUpscale),
+    createDefaultSlice: createDefaultSlice(defaultParams),
+    createHistorySlice: createHistorySlice(),
+    createImg2ImgSlice: createImg2ImgSlice(server, defaultParams, defaultHighres, defaultModel, defaultUpscale),
+    createInpaintSlice: createInpaintSlice(server, defaultParams, defaultHighres, defaultModel, defaultUpscale),
+    createModelSlice: createModelSlice(),
+    createProfileSlice: createProfileSlice(),
+    createResetSlice: createResetSlice(),
+    createTxt2ImgSlice: createTxt2ImgSlice(server, defaultParams, defaultHighres, defaultModel, defaultUpscale, defaultGrid),
+    createUpscaleSlice: createUpscaleSlice(defaultParams, defaultHighres, defaultModel, defaultUpscale),
+  };
+}
diff --git a/gui/src/state/history.ts b/gui/src/state/history.ts
new file mode 100644
index 00000000..de71ef72
--- /dev/null
+++ b/gui/src/state/history.ts
@@ -0,0 +1,68 @@
+import { Maybe } from '@apextoaster/js-utils';
+import { ImageResponse, ReadyResponse, RetryParams } from '../types/api.js';
+import { Slice } from './types.js';
+import { DEFAULT_HISTORY } from '../constants.js';
+
+export interface HistoryItem {
+  image: ImageResponse;
+  ready: Maybe<ReadyResponse>;
+  retry: Maybe<RetryParams>;
+}
+
+export interface HistorySlice {
+  history: Array<HistoryItem>;
+  limit: number;
+
+  pushHistory(image: ImageResponse, retry?: RetryParams): void;
+  removeHistory(image: ImageResponse): void;
+  setLimit(limit: number): void;
+  setReady(image: ImageResponse, ready: ReadyResponse): void;
+}
+
+export function createHistorySlice<TState extends HistorySlice>(): Slice<TState, HistorySlice> {
+  return (set) => ({
+    history: [],
+    limit: DEFAULT_HISTORY.limit,
+    pushHistory(image, retry) {
+      set((prev) => ({
+        ...prev,
+        history: [
+          {
+            image,
+            ready: undefined,
+            retry,
+          },
+          ...prev.history,
+        ].slice(0, prev.limit + DEFAULT_HISTORY.scrollback),
+      }));
+    },
+    removeHistory(image) {
+      set((prev) => ({
+        ...prev,
+        history: prev.history.filter((it) => it.image.outputs[0].key !== image.outputs[0].key),
+      }));
+    },
+    setLimit(limit) {
+      set((prev) => ({
+        ...prev,
+        limit,
+      }));
+    },
+    setReady(image, ready) {
+      set((prev) => {
+        const history = [...prev.history];
+        const idx = history.findIndex((it) => it.image.outputs[0].key === image.outputs[0].key);
+        if (idx >= 0) {
+          history[idx].ready = ready;
+        } else {
+          // TODO: error
+        }
+
+        return {
+          ...prev,
+          history,
+        };
+      });
+    },
+  });
+}
diff --git a/gui/src/state/img2img.ts b/gui/src/state/img2img.ts
new file mode 100644
index 00000000..b8f986b0
--- /dev/null
+++ b/gui/src/state/img2img.ts
@@ -0,0 +1,98 @@
+
+import { ServerParams } from '../config.js';
+import {
+  BaseImgParams,
+  HighresParams,
+  Img2ImgParams,
+  ModelParams,
+  UpscaleParams,
+} from '../types/params.js';
+import { Slice, TabState } from './types.js';
+
+export interface Img2ImgSlice {
+  img2img: TabState<Img2ImgParams>;
+  img2imgModel: ModelParams;
+  img2imgHighres: HighresParams;
+  img2imgUpscale: UpscaleParams;
+
+  resetImg2Img(): void;
+
+  setImg2Img(params: Partial<Img2ImgParams>): void;
+  setImg2ImgModel(params: Partial<ModelParams>): void;
+  setImg2ImgHighres(params: Partial<HighresParams>): void;
+  setImg2ImgUpscale(params: Partial<UpscaleParams>): void;
+}
+
+// eslint-disable-next-line max-params
+export function createImg2ImgSlice<TState extends Img2ImgSlice>(
+  server: ServerParams,
+  defaultParams: Required<BaseImgParams>,
+  defaultHighres: HighresParams,
+  defaultModel: ModelParams,
+  defaultUpscale: UpscaleParams
+): Slice<TState, Img2ImgSlice> {
+  return (set) => ({
+    img2img: {
+      ...defaultParams,
+      loopback: server.loopback.default,
+      // eslint-disable-next-line no-null/no-null
+      source: null,
+      sourceFilter: '',
+      strength: server.strength.default,
+    },
+    img2imgHighres: {
+      ...defaultHighres,
+    },
+    img2imgModel: {
+      ...defaultModel,
+    },
+    img2imgUpscale: {
+      ...defaultUpscale,
+    },
+    resetImg2Img() {
+      set({
+        img2img: {
+          ...defaultParams,
+          loopback: server.loopback.default,
+          // eslint-disable-next-line no-null/no-null
+          source: null,
+          sourceFilter: '',
+          strength: server.strength.default,
+        },
+      } as Partial<TState>);
+    },
+    setImg2Img(params) {
+      set((prev) => ({
+        img2img: {
+          ...prev.img2img,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setImg2ImgHighres(params) {
+      set((prev) => ({
+        img2imgHighres: {
+          ...prev.img2imgHighres,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setImg2ImgModel(params) {
+      set((prev) => ({
+        img2imgModel: {
+          ...prev.img2imgModel,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setImg2ImgUpscale(params) {
+      set((prev) => ({
+        img2imgUpscale: {
+          ...prev.img2imgUpscale,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+  });
+
+}
diff --git a/gui/src/state/inpaint.ts b/gui/src/state/inpaint.ts
new file mode 100644
index 00000000..7dab5af3
--- /dev/null
+++ b/gui/src/state/inpaint.ts
@@ -0,0 +1,136 @@
+import { ServerParams } from '../config.js';
+import { DEFAULT_BRUSH } from '../constants.js';
+import {
+  BaseImgParams,
+  BrushParams,
+  HighresParams,
+  InpaintParams,
+  ModelParams,
+  OutpaintPixels,
+  UpscaleParams,
+} from '../types/params.js';
+import { Slice, TabState } from './types.js';
+export interface InpaintSlice {
+  inpaint: TabState<InpaintParams>;
+  inpaintBrush: BrushParams;
+  inpaintModel: ModelParams;
+  inpaintHighres: HighresParams;
+  inpaintUpscale: UpscaleParams;
+  outpaint: OutpaintPixels;
+
+  resetInpaint(): void;
+
+  setInpaint(params: Partial<InpaintParams>): void;
+  setInpaintBrush(brush: Partial<BrushParams>): void;
+  setInpaintModel(params: Partial<ModelParams>): void;
+  setInpaintHighres(params: Partial<HighresParams>): void;
+  setInpaintUpscale(params: Partial<UpscaleParams>): void;
+  setOutpaint(pixels: Partial<OutpaintPixels>): void;
+}
+
+// eslint-disable-next-line max-params
+export function createInpaintSlice<TState extends InpaintSlice>(
+  server: ServerParams,
+  defaultParams: Required<BaseImgParams>,
+  defaultHighres: HighresParams,
+  defaultModel: ModelParams,
+  defaultUpscale: UpscaleParams,
+): Slice<TState, InpaintSlice> {
+  return (set) => ({
+    inpaint: {
+      ...defaultParams,
+      fillColor: server.fillColor.default,
+      filter: server.filter.default,
+      // eslint-disable-next-line no-null/no-null
+      mask: null,
+      noise: server.noise.default,
+      // eslint-disable-next-line no-null/no-null
+      source: null,
+      strength: server.strength.default,
+      tileOrder: server.tileOrder.default,
+    },
+    inpaintBrush: {
+      ...DEFAULT_BRUSH,
+    },
+    inpaintHighres: {
+      ...defaultHighres,
+    },
+    inpaintModel: {
+      ...defaultModel,
+    },
+    inpaintUpscale: {
+      ...defaultUpscale,
+    },
+    outpaint: {
+      enabled: false,
+      left: server.left.default,
+      right: server.right.default,
+      top: server.top.default,
+      bottom: server.bottom.default,
+    },
+    resetInpaint() {
+      set({
+        inpaint: {
+          ...defaultParams,
+          fillColor: server.fillColor.default,
+          filter: server.filter.default,
+          // eslint-disable-next-line no-null/no-null
+          mask: null,
+          noise: server.noise.default,
+          // eslint-disable-next-line no-null/no-null
+          source: null,
+          strength: server.strength.default,
+          tileOrder: server.tileOrder.default,
+        },
+      } as Partial<TState>);
+    },
+    setInpaint(params) {
+      set((prev) => ({
+        inpaint: {
+          ...prev.inpaint,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setInpaintBrush(brush) {
+      set((prev) => ({
+        inpaintBrush: {
+          ...prev.inpaintBrush,
+          ...brush,
+        },
+      } as Partial<TState>));
+    },
+    setInpaintHighres(params) {
+      set((prev) => ({
+        inpaintHighres: {
+          ...prev.inpaintHighres,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setInpaintModel(params) {
+      set((prev) => ({
+        inpaintModel: {
+          ...prev.inpaintModel,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setInpaintUpscale(params) {
+      set((prev) => ({
+        inpaintUpscale: {
+          ...prev.inpaintUpscale,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setOutpaint(pixels) {
+      set((prev) => ({
+        outpaint: {
+          ...prev.outpaint,
+          ...pixels,
+        }
+      } as Partial<TState>));
+    },
+  });
+}
diff --git a/gui/src/state/migration/default.ts b/gui/src/state/migration/default.ts
new file mode 100644
index 00000000..65a4353b
--- /dev/null
+++ b/gui/src/state/migration/default.ts
@@ -0,0 +1,93 @@
+/* eslint-disable camelcase */
+import { Logger } from 'browser-bunyan';
+import { ServerParams } from '../../config.js';
+import { BaseImgParams } from '../../types/params.js';
+import { OnnxState, STATE_VERSION } from '../full.js';
+import { Img2ImgSlice } from '../img2img.js';
+import { InpaintSlice } from '../inpaint.js';
+import { Txt2ImgSlice } from '../txt2img.js';
+import { UpscaleSlice } from '../upscale.js';
+
+// #region V7
+export const V7 = 7;
+
+export type BaseImgParamsV7<T extends BaseImgParams> = Omit<T, AddedKeysV11> & {
+  overlap: number;
+  tile: number;
+};
+
+export type OnnxStateV7 = Omit<OnnxState, 'img2img' | 'txt2img'> & {
+  img2img: BaseImgParamsV7<Img2ImgSlice['img2img']>;
+  inpaint: BaseImgParamsV7<InpaintSlice['inpaint']>;
+  txt2img: BaseImgParamsV7<Txt2ImgSlice['txt2img']>;
+  upscale: BaseImgParamsV7<UpscaleSlice['upscale']>;
+};
+// #endregion
+
+// #region V11
+export const REMOVED_KEYS_V11 = ['tile', 'overlap'] as const;
+
+export type RemovedKeysV11 = typeof REMOVED_KEYS_V11[number];
+
+// TODO: can the compiler calculate this?
+export type AddedKeysV11 = 'unet_tile' | 'unet_overlap' | 'vae_tile' | 'vae_overlap';
+// #endregion
+
+// add versions to this list as they are replaced
+export type PreviousState = OnnxStateV7;
+
+// always the latest version
+export type CurrentState = OnnxState;
+
+// any version of state
+export type UnknownState = PreviousState | CurrentState;
+
+export function applyStateMigrations(params: ServerParams, previousState: UnknownState, version: number, logger: Logger): OnnxState {
+  logger.info('applying state migrations from version %s to version %s', version, STATE_VERSION);
+
+  if (version <= V7) {
+    return migrateV7ToV11(params, previousState as PreviousState);
+  }
+
+  return previousState as CurrentState;
+}
+
+export function migrateV7ToV11(params: ServerParams, previousState: PreviousState): CurrentState {
+  // add any missing keys
+  const result: CurrentState = {
+    ...params,
+    ...previousState,
+    img2img: {
+      ...previousState.img2img,
+      unet_overlap: params.unet_overlap.default,
+      unet_tile: params.unet_tile.default,
+      vae_overlap: params.vae_overlap.default,
+      vae_tile: params.vae_tile.default,
+    },
+    inpaint: {
+      ...previousState.inpaint,
+      unet_overlap: params.unet_overlap.default,
+      unet_tile: params.unet_tile.default,
+      vae_overlap: params.vae_overlap.default,
+      vae_tile: params.vae_tile.default,
+    },
+    txt2img: {
+      ...previousState.txt2img,
+      unet_overlap: params.unet_overlap.default,
+      unet_tile: params.unet_tile.default,
+      vae_overlap: params.vae_overlap.default,
+      vae_tile: params.vae_tile.default,
+    },
+    upscale: {
+      ...previousState.upscale,
+      unet_overlap: params.unet_overlap.default,
+      unet_tile: params.unet_tile.default,
+      vae_overlap: params.vae_overlap.default,
+      vae_tile: params.vae_tile.default,
+    },
+  };
+
+  // TODO: remove extra keys
+
+  return result;
+}
diff --git a/gui/src/state/model.ts b/gui/src/state/model.ts
new file mode 100644
index 00000000..3a473182
--- /dev/null
+++ b/gui/src/state/model.ts
@@ -0,0 +1,202 @@
+import { CorrectionModel, DiffusionModel, ExtraNetwork, ExtraSource, ExtrasFile, UpscalingModel } from '../types/model.js';
+import { MISSING_INDEX, Slice } from './types.js';
+
+export interface ModelSlice {
+  extras: ExtrasFile;
+
+  removeCorrectionModel(model: CorrectionModel): void;
+  removeDiffusionModel(model: DiffusionModel): void;
+  removeExtraNetwork(model: ExtraNetwork): void;
+  removeExtraSource(model: ExtraSource): void;
+  removeUpscalingModel(model: UpscalingModel): void;
+
+  setExtras(extras: Partial<ExtrasFile>): void;
+
+  setCorrectionModel(model: CorrectionModel): void;
+  setDiffusionModel(model: DiffusionModel): void;
+  setExtraNetwork(model: ExtraNetwork): void;
+  setExtraSource(model: ExtraSource): void;
+  setUpscalingModel(model: UpscalingModel): void;
+}
+
+// eslint-disable-next-line sonarjs/cognitive-complexity
+export function createModelSlice<TState extends ModelSlice>(): Slice<TState, ModelSlice> {
+  // eslint-disable-next-line sonarjs/cognitive-complexity
+  return (set) => ({
+    extras: {
+      correction: [],
+      diffusion: [],
+      networks: [],
+      sources: [],
+      upscaling: [],
+    },
+    setExtras(extras) {
+      set((prev) => ({
+        ...prev,
+        extras: {
+          ...prev.extras,
+          ...extras,
+        },
+      }));
+    },
+    setCorrectionModel(model) {
+      set((prev) => {
+        const correction = [...prev.extras.correction];
+        const exists = correction.findIndex((it) => model.name === it.name);
+        if (exists === MISSING_INDEX) {
+          correction.push(model);
+        } else {
+          correction[exists] = model;
+        }
+
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            correction,
+          },
+        };
+      });
+    },
+    setDiffusionModel(model) {
+      set((prev) => {
+        const diffusion = [...prev.extras.diffusion];
+        const exists = diffusion.findIndex((it) => model.name === it.name);
+        if (exists === MISSING_INDEX) {
+          diffusion.push(model);
+        } else {
+          diffusion[exists] = model;
+        }
+
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            diffusion,
+          },
+        };
+      });
+    },
+    setExtraNetwork(model) {
+      set((prev) => {
+        const networks = [...prev.extras.networks];
+        const exists = networks.findIndex((it) => model.name === it.name);
+        if (exists === MISSING_INDEX) {
+          networks.push(model);
+        } else {
+          networks[exists] = model;
+        }
+
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            networks,
+          },
+        };
+      });
+    },
+    setExtraSource(model) {
+      set((prev) => {
+        const sources = [...prev.extras.sources];
+        const exists = sources.findIndex((it) => model.name === it.name);
+        if (exists === MISSING_INDEX) {
+          sources.push(model);
+        } else {
+          sources[exists] = model;
+        }
+
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            sources,
+          },
+        };
+      });
+    },
+    setUpscalingModel(model) {
+      set((prev) => {
+        const upscaling = [...prev.extras.upscaling];
+        const exists = upscaling.findIndex((it) => model.name === it.name);
+        if (exists === MISSING_INDEX) {
+          upscaling.push(model);
+        } else {
+          upscaling[exists] = model;
+        }
+
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            upscaling,
+          },
+        };
+      });
+    },
+    removeCorrectionModel(model) {
+      set((prev) => {
+        const correction = prev.extras.correction.filter((it) => model.name !== it.name);;
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            correction,
+          },
+        };
+      });
+
+    },
+    removeDiffusionModel(model) {
+      set((prev) => {
+        const diffusion = prev.extras.diffusion.filter((it) => model.name !== it.name);;
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            diffusion,
+          },
+        };
+      });
+
+    },
+    removeExtraNetwork(model) {
+      set((prev) => {
+        const networks = prev.extras.networks.filter((it) => model.name !== it.name);;
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            networks,
+          },
+        };
+      });
+
+    },
+    removeExtraSource(model) {
+      set((prev) => {
+        const sources = prev.extras.sources.filter((it) => model.name !== it.name);;
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            sources,
+          },
+        };
+      });
+
+    },
+    removeUpscalingModel(model) {
+      set((prev) => {
+        const upscaling = prev.extras.upscaling.filter((it) => model.name !== it.name);;
+        return {
+          ...prev,
+          extras: {
+            ...prev.extras,
+            upscaling,
+          },
+        };
+      });
+    },
+  });
+}
diff --git a/gui/src/state/profile.ts b/gui/src/state/profile.ts
new file mode 100644
index 00000000..73d52eab
--- /dev/null
+++ b/gui/src/state/profile.ts
@@ -0,0 +1,52 @@
+import { Maybe } from '@apextoaster/js-utils';
+import { BaseImgParams, HighresParams, Txt2ImgParams, UpscaleParams } from '../types/params.js';
+import { Slice } from './types.js';
+
+export interface ProfileItem {
+  name: string;
+  params: BaseImgParams | Txt2ImgParams;
+  highres?: Maybe<HighresParams>;
+  upscale?: Maybe<UpscaleParams>;
+}
+
+export interface ProfileSlice {
+  profiles: Array<ProfileItem>;
+
+  removeProfile(profileName: string): void;
+
+  saveProfile(profile: ProfileItem): void;
+}
+
+export function createProfileSlice<TState extends ProfileSlice>(): Slice<TState, ProfileSlice> {
+  return (set) => ({
+    profiles: [],
+    saveProfile(profile: ProfileItem) {
+      set((prev) => {
+        const profiles = [...prev.profiles];
+        const idx = profiles.findIndex((it) => it.name === profile.name);
+        if (idx >= 0) {
+          profiles[idx] = profile;
+        } else {
+          profiles.push(profile);
+        }
+        return {
+          ...prev,
+          profiles,
+        };
+      });
+    },
+    removeProfile(profileName: string) {
+      set((prev) => {
+        const profiles = [...prev.profiles];
+        const idx = profiles.findIndex((it) => it.name === profileName);
+        if (idx >= 0) {
+          profiles.splice(idx, 1);
+        }
+        return {
+          ...prev,
+          profiles,
+        };
+      });
+    }
+  });
+}
diff --git a/gui/src/state/reset.ts b/gui/src/state/reset.ts
new file mode 100644
index 00000000..53272e5c
--- /dev/null
+++ b/gui/src/state/reset.ts
@@ -0,0 +1,28 @@
+import { BlendSlice } from './blend.js';
+import { Img2ImgSlice } from './img2img.js';
+import { InpaintSlice } from './inpaint.js';
+import { Txt2ImgSlice } from './txt2img.js';
+import { Slice } from './types.js';
+import { UpscaleSlice } from './upscale.js';
+
+export type SlicesWithReset = Txt2ImgSlice & Img2ImgSlice & InpaintSlice & UpscaleSlice & BlendSlice;
+
+export interface ResetSlice {
+  resetAll(): void;
+}
+
+export function createResetSlice<TState extends ResetSlice & SlicesWithReset>(): Slice<TState, ResetSlice> {
+  return (set) => ({
+    resetAll() {
+      set((prev) => {
+        const next = { ...prev };
+        next.resetImg2Img();
+        next.resetInpaint();
+        next.resetTxt2Img();
+        next.resetUpscale();
+        next.resetBlend();
+        return next;
+      });
+    },
+  });
+}
diff --git a/gui/src/state/txt2img.ts b/gui/src/state/txt2img.ts
new file mode 100644
index 00000000..8bec3273
--- /dev/null
+++ b/gui/src/state/txt2img.ts
@@ -0,0 +1,105 @@
+import { PipelineGrid } from '../client/utils.js';
+import { ServerParams } from '../config.js';
+import {
+  BaseImgParams,
+  HighresParams,
+  ModelParams,
+  Txt2ImgParams,
+  UpscaleParams,
+} from '../types/params.js';
+import { Slice, TabState } from './types.js';
+
+export interface Txt2ImgSlice {
+  txt2img: TabState<Txt2ImgParams>;
+  txt2imgModel: ModelParams;
+  txt2imgHighres: HighresParams;
+  txt2imgUpscale: UpscaleParams;
+  txt2imgVariable: PipelineGrid;
+
+  resetTxt2Img(): void;
+
+  setTxt2Img(params: Partial<Txt2ImgParams>): void;
+  setTxt2ImgModel(params: Partial<ModelParams>): void;
+  setTxt2ImgHighres(params: Partial<HighresParams>): void;
+  setTxt2ImgUpscale(params: Partial<UpscaleParams>): void;
+  setTxt2ImgVariable(params: Partial<PipelineGrid>): void;
+}
+
+// eslint-disable-next-line max-params
+export function createTxt2ImgSlice<TState extends Txt2ImgSlice>(
+  server: ServerParams,
+  defaultParams: Required<BaseImgParams>,
+  defaultHighres: HighresParams,
+  defaultModel: ModelParams,
+  defaultUpscale: UpscaleParams,
+  defaultGrid: PipelineGrid,
+): Slice<TState, Txt2ImgSlice> {
+  return (set) => ({
+    txt2img: {
+      ...defaultParams,
+      width: server.width.default,
+      height: server.height.default,
+    },
+    txt2imgHighres: {
+      ...defaultHighres,
+    },
+    txt2imgModel: {
+      ...defaultModel,
+    },
+    txt2imgUpscale: {
+      ...defaultUpscale,
+    },
+    txt2imgVariable: {
+      ...defaultGrid,
+    },
+    setTxt2Img(params) {
+      set((prev) => ({
+        txt2img: {
+          ...prev.txt2img,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setTxt2ImgHighres(params) {
+      set((prev) => ({
+        txt2imgHighres: {
+          ...prev.txt2imgHighres,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setTxt2ImgModel(params) {
+      set((prev) => ({
+        txt2imgModel: {
+          ...prev.txt2imgModel,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setTxt2ImgUpscale(params) {
+      set((prev) => ({
+        txt2imgUpscale: {
+          ...prev.txt2imgUpscale,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setTxt2ImgVariable(params) {
+      set((prev) => ({
+        txt2imgVariable: {
+          ...prev.txt2imgVariable,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    resetTxt2Img() {
+      set({
+        txt2img: {
+          ...defaultParams,
+          width: server.width.default,
+          height: server.height.default,
+        },
+      } as Partial<TState>);
+    },
+  });
+}
diff --git a/gui/src/state/types.ts b/gui/src/state/types.ts
new file mode 100644
index 00000000..b1e9d3ad
--- /dev/null
+++ b/gui/src/state/types.ts
@@ -0,0 +1,17 @@
+import { PaletteMode } from '@mui/material';
+import { StateCreator } from 'zustand';
+import { ConfigFiles, ConfigState } from '../config.js';
+
+export const MISSING_INDEX = -1;
+
+export type Theme = PaletteMode | ''; // tri-state, '' is unset
+
+/**
+ * Combine optional files and required ranges.
+ */
+export type TabState<TabParams> = ConfigFiles<Required<TabParams>> & ConfigState<Required<TabParams>>;
+
+/**
+ * Shorthand for state creator to reduce repeated arguments.
+ */
+export type Slice<TState, TValue> = StateCreator<TState, [], [], TValue>;
diff --git a/gui/src/state/upscale.ts b/gui/src/state/upscale.ts
new file mode 100644
index 00000000..e78d689a
--- /dev/null
+++ b/gui/src/state/upscale.ts
@@ -0,0 +1,87 @@
+import {
+  BaseImgParams,
+  HighresParams,
+  ModelParams,
+  UpscaleParams,
+  UpscaleReqParams,
+} from '../types/params.js';
+import { Slice, TabState } from './types.js';
+
+export interface UpscaleSlice {
+  upscale: TabState<UpscaleReqParams>;
+  upscaleHighres: HighresParams;
+  upscaleModel: ModelParams;
+  upscaleUpscale: UpscaleParams;
+
+  resetUpscale(): void;
+
+  setUpscale(params: Partial<UpscaleReqParams>): void;
+  setUpscaleHighres(params: Partial<HighresParams>): void;
+  setUpscaleModel(params: Partial<ModelParams>): void;
+  setUpscaleUpscale(params: Partial<UpscaleParams>): void;
+}
+
+export function createUpscaleSlice<TState extends UpscaleSlice>(
+  defaultParams: Required<BaseImgParams>,
+  defaultHighres: HighresParams,
+  defaultModel: ModelParams,
+  defaultUpscale: UpscaleParams,
+): Slice<TState, UpscaleSlice> {
+  return (set) => ({
+    upscale: {
+      ...defaultParams,
+      // eslint-disable-next-line no-null/no-null
+      source: null,
+    },
+    upscaleHighres: {
+      ...defaultHighres,
+    },
+    upscaleModel: {
+      ...defaultModel,
+    },
+    upscaleUpscale: {
+      ...defaultUpscale,
+    },
+    resetUpscale() {
+      set({
+        upscale: {
+          ...defaultParams,
+          // eslint-disable-next-line no-null/no-null
+          source: null,
+        },
+      } as Partial<TState>);
+    },
+    setUpscale(source) {
+      set((prev) => ({
+        upscale: {
+          ...prev.upscale,
+          ...source,
+        },
+      } as Partial<TState>));
+    },
+    setUpscaleHighres(params) {
+      set((prev) => ({
+        upscaleHighres: {
+          ...prev.upscaleHighres,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+    setUpscaleModel(params) {
+      set((prev) => ({
+        upscaleModel: {
+          ...prev.upscaleModel,
+          ...defaultModel,
+        },
+      } as Partial<TState>));
+    },
+    setUpscaleUpscale(params) {
+      set((prev) => ({
+        upscaleUpscale: {
+          ...prev.upscaleUpscale,
+          ...params,
+        },
+      } as Partial<TState>));
+    },
+  });
+}
diff --git a/gui/src/strings/en.ts b/gui/src/strings/en.ts
index c9d901f9..304c3335 100644
--- a/gui/src/strings/en.ts
+++ b/gui/src/strings/en.ts
@@ -287,6 +287,7 @@ export const I18N_STRINGS_EN = {
         'ddpm': 'DDPM',
         'deis-multi': 'DEIS Multistep',
         'dpm-multi': 'DPM Multistep',
+        'dpm-sde': 'DPM SDE (Turbo)',
         'dpm-single': 'DPM Singlestep',
         'euler': 'Euler',
         'euler-a': 'Euler Ancestral',