diff --git a/api/.gitignore b/api/.gitignore
index 17c17360..165cb300 100644
--- a/api/.gitignore
+++ b/api/.gitignore
@@ -2,6 +2,7 @@
 coverage.xml
 entry.py
 
+*.dot
 *.log
 *.swp
 *.pyc
diff --git a/api/onnx_web/prompt/grammar.py b/api/onnx_web/prompt/grammar.py
new file mode 100644
index 00000000..ae28d23e
--- /dev/null
+++ b/api/onnx_web/prompt/grammar.py
@@ -0,0 +1,84 @@
+from typing import List, Union
+
+from arpeggio import EOF, OneOrMore, PTNodeVisitor, RegExMatch
+
+
+def token():
+    return RegExMatch(r"\w+")
+
+
+def token_run():
+    return OneOrMore(token)
+
+
+def phrase_inner():
+    return [phrase, token_run]
+
+
+def pos_phrase():
+    return ("(", OneOrMore(phrase_inner), ")")
+
+
+def neg_phrase():
+    return ("[", OneOrMore(phrase_inner), "]")
+
+
+def phrase():
+    return [pos_phrase, neg_phrase, token_run]
+
+
+def prompt():
+    return OneOrMore(phrase), EOF
+
+
+class PromptPhrase:
+    def __init__(self, tokens: Union[List[str], str], weight: float = 1.0) -> None:
+        self.tokens = tokens
+        self.weight = weight
+
+    def __repr__(self) -> str:
+        return f"{self.tokens} * {self.weight}"
+
+    def __eq__(self, other: object) -> bool:
+        if isinstance(other, self.__class__):
+            return other.tokens == self.tokens and other.weight == self.weight
+
+
+class OnnxPromptVisitor(PTNodeVisitor):
+    def __init__(self, defaults=True, weight=0.5, **kwargs):
+        super().__init__(defaults, **kwargs)
+
+        self.neg_weight = weight
+        self.pos_weight = 1.0 + weight
+
+    def visit_token(self, node, children):
+        return str(node.value)
+
+    def visit_token_run(self, node, children):
+        return children
+
+    def visit_phrase_inner(self, node, children):
+        if isinstance(children[0], PromptPhrase):
+            return children[0]
+        else:
+            return PromptPhrase(children[0])
+
+    def visit_pos_phrase(self, node, children):
+        c = children[0]
+        if isinstance(c, PromptPhrase):
+            return PromptPhrase(c.tokens, c.weight * self.pos_weight)
+        elif isinstance(c, str):
+            return PromptPhrase(c, self.pos_weight)
+
+    def visit_neg_phrase(self, node, children):
+        c = children[0]
+        if isinstance(c, PromptPhrase):
+            return PromptPhrase(c.tokens, c.weight * self.neg_weight)
+        elif isinstance(c, str):
+            return PromptPhrase(c, self.neg_weight)
+
+    def visit_phrase(self, node, children):
+        return children[0]
+
+    def visit_prompt(self, node, children):
+        return children
diff --git a/api/onnx_web/prompt/parser.py b/api/onnx_web/prompt/parser.py
new file mode 100644
index 00000000..6a0fde22
--- /dev/null
+++ b/api/onnx_web/prompt/parser.py
@@ -0,0 +1,47 @@
+from typing import Literal
+
+import numpy as np
+from arpeggio import ParserPython, visit_parse_tree
+
+from .grammar import OnnxPromptVisitor
+from .grammar import prompt as prompt_base
+
+
+def parse_prompt_compel(pipeline, prompt: str) -> np.ndarray:
+    from compel import Compel
+
+    parser = Compel(tokenizer=pipeline.tokenizer, text_encoder=pipeline.text_encoder)
+    return parser([prompt])
+
+
+def parse_prompt_lpw(pipeline, prompt: str, debug=False) -> np.ndarray:
+    pass
+
+
+def parse_prompt_onnx(pipeline, prompt: str, debug=False) -> np.ndarray:
+    parser = ParserPython(prompt_base, debug=debug)
+    visitor = OnnxPromptVisitor()
+
+    ast = parser.parse(prompt)
+    return visit_parse_tree(ast, visitor)
+
+
+def parse_prompt_vanilla(pipeline, prompt: str) -> np.ndarray:
+    return pipeline._encode_prompt(prompt)
+
+
+def parse_prompt(
+    pipeline,
+    prompt: str,
+    engine: Literal["compel", "lpw", "onnx-web", "pipeline"] = "onnx-web",
+) -> np.ndarray:
+    if engine == "compel":
+        return parse_prompt_compel(pipeline, prompt)
+    if engine == "lpw":
+        return parse_prompt_lpw(pipeline, prompt)
+    elif engine == "onnx-web":
+        return parse_prompt_onnx(pipeline, prompt)
+    elif engine == "pipeline":
+        return parse_prompt_vanilla(pipeline, prompt)
+    else:
+        raise ValueError("invalid prompt parser")
diff --git a/api/tests/prompt/__init__.py b/api/tests/prompt/__init__.py
new file mode 100644
index 00000000..e69de29b
diff --git a/api/tests/prompt/test_parser.py b/api/tests/prompt/test_parser.py
new file mode 100644
index 00000000..20c03341
--- /dev/null
+++ b/api/tests/prompt/test_parser.py
@@ -0,0 +1,37 @@
+import unittest
+from onnx_web.prompt.grammar import PromptPhrase
+from onnx_web.prompt.parser import parse_prompt_onnx
+
+class ParserTests(unittest.TestCase):
+    def test_single_word_phrase(self):
+        res = parse_prompt_onnx(None, "foo (bar) bin", debug=False)
+        self.assertListEqual(
+            [str(i) for i in res],
+            [
+                str(["foo"]),
+                str(PromptPhrase(["bar"], weight=1.5)),
+                str(["bin"]),
+            ]
+        )
+
+    def test_multi_word_phrase(self):
+        res = parse_prompt_onnx(None, "foo bar (middle words) bin bun", debug=False)
+        self.assertListEqual(
+            [str(i) for i in res],
+            [
+                str(["foo", "bar"]),
+                str(PromptPhrase(["middle", "words"], weight=1.5)),
+                str(["bin", "bun"]),
+            ]
+        )
+
+    def test_nested_phrase(self):
+        res = parse_prompt_onnx(None, "foo (((bar))) bin", debug=False)
+        self.assertListEqual(
+            [str(i) for i in res],
+            [
+                str(["foo"]),
+                str(PromptPhrase(["bar"], weight=(1.5 ** 3))),
+                str(["bin"]),
+            ]
+        )