diff --git a/daras_ai_v2/asr.py b/daras_ai_v2/asr.py
index 699e9fe2a..44a05749d 100644
--- a/daras_ai_v2/asr.py
+++ b/daras_ai_v2/asr.py
@@ -3,7 +3,6 @@
 import os.path
 import tempfile
 import typing
-from enum import Enum
 
 import gooey_gui as gui
 import requests
@@ -14,6 +13,7 @@
 from daras_ai.image_input import upload_file_from_bytes, gs_url_to_uri
 from daras_ai_v2 import settings
 from daras_ai_v2.azure_asr import azure_asr
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.exceptions import (
     raise_for_status,
     UserError,
@@ -204,7 +204,7 @@
 GHANA_NLP_MAXLEN = 500
 
 
-class AsrModels(Enum):
+class AsrModels(GooeyEnum):
     whisper_large_v2 = "Whisper Large v2 (openai)"
     whisper_large_v3 = "Whisper Large v3 (openai)"
     whisper_hindi_large_v2 = "Whisper Hindi Large v2 (Bhashini)"
@@ -277,7 +277,7 @@ class AsrOutputJson(typing_extensions.TypedDict):
     chunks: typing_extensions.NotRequired[list[AsrChunk]]
 
 
-class AsrOutputFormat(Enum):
+class AsrOutputFormat(GooeyEnum):
     text = "Text"
     json = "JSON"
     srt = "SRT"
@@ -290,7 +290,7 @@ class TranslationModel(typing.NamedTuple):
     supports_auto_detect: bool = False
 
 
-class TranslationModels(TranslationModel, Enum):
+class TranslationModels(TranslationModel, GooeyEnum):
     google = TranslationModel(
         label="Google Translate",
         supports_glossary=True,
diff --git a/daras_ai_v2/base.py b/daras_ai_v2/base.py
index 70c294a92..0874ba64f 100644
--- a/daras_ai_v2/base.py
+++ b/daras_ai_v2/base.py
@@ -127,6 +127,7 @@ class BasePage:
     title: str
     workflow: Workflow
     slug_versions: list[str]
+    sdk_method_name: str
 
     sane_defaults: dict = {}
 
@@ -144,6 +145,9 @@ class BasePage:
     )
 
     class RequestModel(BaseModel):
+        class Config:
+            use_enum_values = True
+
         functions: list[RecipeFunction] | None = Field(
             title="🧩 Developer Tools and Functions",
         )
@@ -317,6 +321,12 @@ def sentry_event_set_user(self, event, hint):
             }
         return event
 
+    @classmethod
+    def get_openapi_extra(cls) -> dict[str, typing.Any]:
+        return {
+            "x-fern-sdk-method-name": cls.sdk_method_name,
+        }
+
     def refresh_state(self):
         sr = self.current_sr
         channel = self.realtime_channel_name(sr.run_id, sr.uid)
diff --git a/daras_ai_v2/custom_enum.py b/daras_ai_v2/custom_enum.py
index b9aacb843..270d4834c 100644
--- a/daras_ai_v2/custom_enum.py
+++ b/daras_ai_v2/custom_enum.py
@@ -1,12 +1,38 @@
+import functools
 import typing
 from enum import Enum
 
 import typing_extensions
 
+
+def cached_classmethod(func: typing.Callable):
+    """
+    This cache is a hack to get around a bug where
+    dynamic Enums with the same name cause a crash
+    when generating the OpenAPI spec.
+    """
+
+    @functools.wraps(func)
+    def wrapper(cls):
+        if not hasattr(cls, "_cached_classmethod"):
+            cls._cached_classmethod = {}
+        if id(func) not in cls._cached_classmethod:
+            cls._cached_classmethod[id(func)] = func(cls)
+
+        return cls._cached_classmethod[id(func)]
+
+    return wrapper
+
+
 T = typing.TypeVar("T", bound="GooeyEnum")
 
 
 class GooeyEnum(Enum):
+    @property
+    def api_value(self):
+        # api_value is usually the name
+        return self.name
+
     @classmethod
     def db_choices(cls):
         return [(e.db_value, e.label) for e in cls]
@@ -20,12 +46,23 @@ def from_db(cls, db_value) -> typing_extensions.Self:
 
     @classmethod
     @property
-    def api_choices(cls):
-        return typing.Literal[tuple(e.name for e in cls)]
+    @cached_classmethod
+    def api_enum(cls):
+        """
+        Enum that is useful as a type in API requests.
+
+        Maps `api_value`->`api_value` (default: `name`->`name`)
+        for all values.
+
+        The title (same as the Enum class name) will be
+        used as the new Enum's title. This will be passed
+        on to the OpenAPI schema and the generated SDK.
+        """
+        return Enum(cls.__name__, {e.api_value: e.api_value for e in cls})
 
     @classmethod
-    def from_api(cls, name: str) -> typing_extensions.Self:
+    def from_api(cls, api_value: str) -> typing_extensions.Self:
         for e in cls:
-            if e.name == name:
+            if e.api_value == api_value:
                 return e
-        raise ValueError(f"Invalid {cls.__name__} {name=}")
+        raise ValueError(f"Invalid {cls.__name__} {api_value=}")
diff --git a/daras_ai_v2/embedding_model.py b/daras_ai_v2/embedding_model.py
index 9f5f3ae1b..5f2927bcc 100644
--- a/daras_ai_v2/embedding_model.py
+++ b/daras_ai_v2/embedding_model.py
@@ -1,7 +1,6 @@
 import hashlib
 import io
 import typing
-from enum import Enum
 from functools import partial
 
 import numpy as np
@@ -13,6 +12,7 @@
 from jinja2.lexer import whitespace_re
 from loguru import logger
 
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.gpu_server import call_celery_task
 from daras_ai_v2.language_model import get_openai_client
 from daras_ai_v2.redis_cache import (
@@ -25,7 +25,7 @@ class EmbeddingModel(typing.NamedTuple):
     label: str
 
 
-class EmbeddingModels(Enum):
+class EmbeddingModels(EmbeddingModel, GooeyEnum):
     openai_3_large = EmbeddingModel(
         model_id=("openai-text-embedding-3-large-prod-ca-1", "text-embedding-3-large"),
         label="Text Embedding 3 Large (OpenAI)",
@@ -65,12 +65,8 @@ class EmbeddingModels(Enum):
     )
 
     @property
-    def model_id(self) -> typing.Iterable[str] | str:
-        return self.value.model_id
-
-    @property
-    def label(self) -> str:
-        return self.value.label
+    def db_value(self):
+        return self.name
 
     @classmethod
     def get(cls, key, default=None):
diff --git a/daras_ai_v2/image_segmentation.py b/daras_ai_v2/image_segmentation.py
index 099832979..004264afc 100644
--- a/daras_ai_v2/image_segmentation.py
+++ b/daras_ai_v2/image_segmentation.py
@@ -1,14 +1,13 @@
-from enum import Enum
-
 import requests
 
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.exceptions import raise_for_status
 from daras_ai_v2.gpu_server import (
     call_celery_task_outfile,
 )
 
 
-class ImageSegmentationModels(Enum):
+class ImageSegmentationModels(str, GooeyEnum):
     dis = "Dichotomous Image Segmentation"
     u2net = "U²-Net"
 
diff --git a/daras_ai_v2/img_model_settings_widgets.py b/daras_ai_v2/img_model_settings_widgets.py
index 7f62fb878..3b9705479 100644
--- a/daras_ai_v2/img_model_settings_widgets.py
+++ b/daras_ai_v2/img_model_settings_widgets.py
@@ -2,11 +2,10 @@
 
 from daras_ai_v2.enum_selector_widget import enum_selector, enum_multiselect
 from daras_ai_v2.stable_diffusion import (
-    Text2ImgModels,
+    TextToImageModels,
     InpaintingModels,
-    Img2ImgModels,
+    ImageToImageModels,
     ControlNetModels,
-    controlnet_model_explanations,
     Schedulers,
 )
 
@@ -37,10 +36,10 @@ def img_model_settings(
     negative_prompt_setting(selected_model)
 
     num_outputs_setting(selected_model)
-    if models_enum is not Img2ImgModels:
+    if models_enum is not ImageToImageModels:
         output_resolution_setting()
 
-    if models_enum is Text2ImgModels:
+    if models_enum is TextToImageModels:
         sd_2_upscaling_setting()
 
     col1, col2 = gui.columns(2)
@@ -49,11 +48,11 @@ def img_model_settings(
         guidance_scale_setting(selected_model)
 
     with col2:
-        if models_enum is Img2ImgModels and not gui.session_state.get(
+        if models_enum is ImageToImageModels and not gui.session_state.get(
             "selected_controlnet_model"
         ):
             prompt_strength_setting(selected_model)
-        if selected_model == Img2ImgModels.instruct_pix2pix.name:
+        if selected_model == ImageToImageModels.instruct_pix2pix.name:
             instruct_pix2pix_settings()
 
     if show_scheduler:
@@ -73,10 +72,10 @@ def model_selector(
     high_explanation: str = "At {high} the control nets will be applied tightly to the prompted visual, possibly overriding the prompt",
 ):
     controlnet_unsupported_models = [
-        Img2ImgModels.instruct_pix2pix.name,
-        Img2ImgModels.dall_e.name,
-        Img2ImgModels.jack_qiao.name,
-        Img2ImgModels.sd_2.name,
+        ImageToImageModels.instruct_pix2pix.name,
+        ImageToImageModels.dall_e.name,
+        ImageToImageModels.jack_qiao.name,
+        ImageToImageModels.sd_2.name,
     ]
     col1, col2 = gui.columns(2)
     with col1:
@@ -96,12 +95,12 @@ def model_selector(
             """
         )
         if (
-            models_enum is Img2ImgModels
+            models_enum is ImageToImageModels
             and gui.session_state.get("selected_model") in controlnet_unsupported_models
         ):
             if "selected_controlnet_model" in gui.session_state:
                 gui.session_state["selected_controlnet_model"] = None
-        elif models_enum is Img2ImgModels:
+        elif models_enum is ImageToImageModels:
             enum_multiselect(
                 ControlNetModels,
                 label=controlnet_explanation,
@@ -130,9 +129,7 @@ def controlnet_settings(
     if not models:
         return
 
-    if extra_explanations is None:
-        extra_explanations = {}
-    explanations = controlnet_model_explanations | extra_explanations
+    extra_explanations = extra_explanations or {}
 
     state_values = gui.session_state.get("controlnet_conditioning_scale", [])
     new_values = []
@@ -157,7 +154,9 @@ def controlnet_settings(
             pass
         new_values.append(
             controlnet_weight_setting(
-                selected_controlnet_model=model, explanations=explanations, key=key
+                selected_controlnet_model=model,
+                extra_explanations=extra_explanations,
+                key=key,
             ),
         )
     gui.session_state["controlnet_conditioning_scale"] = new_values
@@ -166,13 +165,13 @@ def controlnet_settings(
 def controlnet_weight_setting(
     *,
     selected_controlnet_model: str,
-    explanations: dict[ControlNetModels, str],
+    extra_explanations: dict[ControlNetModels, str],
     key: str = "controlnet_conditioning_scale",
 ):
     model = ControlNetModels[selected_controlnet_model]
     return gui.slider(
         label=f"""
-        {explanations[model]}.
+        {extra_explanations.get(model, model.explanation)}.
         """,
         key=key,
         min_value=CONTROLNET_CONDITIONING_SCALE_RANGE[0],
@@ -215,7 +214,7 @@ def quality_setting(selected_models=None):
         return
     if any(
         [
-            selected_model in [Text2ImgModels.dall_e_3.name]
+            selected_model in [TextToImageModels.dall_e_3.name]
             for selected_model in selected_models
         ]
     ):
@@ -375,8 +374,8 @@ def sd_2_upscaling_setting():
 
 def scheduler_setting(selected_model: str = None):
     if selected_model in [
-        Text2ImgModels.dall_e.name,
-        Text2ImgModels.jack_qiao,
+        TextToImageModels.dall_e.name,
+        TextToImageModels.jack_qiao,
     ]:
         return
     enum_selector(
@@ -395,8 +394,8 @@ def scheduler_setting(selected_model: str = None):
 
 def guidance_scale_setting(selected_model: str = None):
     if selected_model in [
-        Text2ImgModels.dall_e.name,
-        Text2ImgModels.jack_qiao,
+        TextToImageModels.dall_e.name,
+        TextToImageModels.jack_qiao,
     ]:
         return
     gui.slider(
@@ -435,8 +434,8 @@ def instruct_pix2pix_settings():
 
 def prompt_strength_setting(selected_model: str = None):
     if selected_model in [
-        Img2ImgModels.dall_e.name,
-        Img2ImgModels.instruct_pix2pix.name,
+        ImageToImageModels.dall_e.name,
+        ImageToImageModels.instruct_pix2pix.name,
     ]:
         return
 
@@ -458,7 +457,7 @@ def prompt_strength_setting(selected_model: str = None):
 
 
 def negative_prompt_setting(selected_model: str = None):
-    if selected_model in [Text2ImgModels.dall_e.name]:
+    if selected_model in [TextToImageModels.dall_e.name]:
         return
 
     gui.text_area(
diff --git a/daras_ai_v2/language_model.py b/daras_ai_v2/language_model.py
index 27f9c5786..6ab0be2e9 100644
--- a/daras_ai_v2/language_model.py
+++ b/daras_ai_v2/language_model.py
@@ -25,6 +25,7 @@
 
 from daras_ai.image_input import gs_url_to_uri, bytes_to_cv2_img, cv2_img_to_bytes
 from daras_ai_v2.asr import get_google_auth_session
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.exceptions import raise_for_status, UserError
 from daras_ai_v2.gpu_server import call_celery_task
 from daras_ai_v2.text_splitter import (
@@ -72,7 +73,7 @@ class LLMSpec(typing.NamedTuple):
     supports_json: bool = False
 
 
-class LargeLanguageModels(Enum):
+class LargeLanguageModels(GooeyEnum):
     # https://platform.openai.com/docs/models/gpt-4o
     gpt_4_o = LLMSpec(
         label="GPT-4o (openai)",
@@ -474,7 +475,9 @@ def get_entry_text(entry: ConversationEntry) -> str:
     )
 
 
-ResponseFormatType = typing.Literal["text", "json_object"]
+class ResponseFormatType(str, GooeyEnum):
+    text = "text"
+    json_object = "json_object"
 
 
 def run_language_model(
diff --git a/daras_ai_v2/lipsync_api.py b/daras_ai_v2/lipsync_api.py
index 7aa81cc12..34f6c7f75 100644
--- a/daras_ai_v2/lipsync_api.py
+++ b/daras_ai_v2/lipsync_api.py
@@ -1,15 +1,15 @@
 import typing
-from enum import Enum
 
 from loguru import logger
 from pydantic import BaseModel, Field
 
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.exceptions import UserError, GPUError
 from daras_ai_v2.gpu_server import call_celery_task_outfile_with_ret
 from daras_ai_v2.pydantic_validation import FieldHttpUrl
 
 
-class LipsyncModel(Enum):
+class LipsyncModels(GooeyEnum):
     Wav2Lip = "SD, Low-res (~480p), Fast (Rudrabha/Wav2Lip)"
     SadTalker = "HD, Hi-res (max 1080p), Slow (OpenTalker/SadTalker)"
 
diff --git a/daras_ai_v2/lipsync_settings_widgets.py b/daras_ai_v2/lipsync_settings_widgets.py
index 515be000b..b06032686 100644
--- a/daras_ai_v2/lipsync_settings_widgets.py
+++ b/daras_ai_v2/lipsync_settings_widgets.py
@@ -1,15 +1,15 @@
 import gooey_gui as gui
 from daras_ai_v2.field_render import field_label_val
 
-from daras_ai_v2.lipsync_api import LipsyncModel, SadTalkerSettings
+from daras_ai_v2.lipsync_api import LipsyncModels, SadTalkerSettings
 
 
 def lipsync_settings(selected_model: str):
     match selected_model:
-        case LipsyncModel.Wav2Lip.name:
+        case LipsyncModels.Wav2Lip.name:
             wav2lip_settings()
             gui.session_state.pop("sadtalker_settings", None)
-        case LipsyncModel.SadTalker.name:
+        case LipsyncModels.SadTalker.name:
             settings = SadTalkerSettings.parse_obj(
                 gui.session_state.setdefault(
                     "sadtalker_settings", SadTalkerSettings().dict()
diff --git a/daras_ai_v2/search_ref.py b/daras_ai_v2/search_ref.py
index 0ed45386e..6ea91368e 100644
--- a/daras_ai_v2/search_ref.py
+++ b/daras_ai_v2/search_ref.py
@@ -1,10 +1,10 @@
 import re
 import typing
-from enum import Enum
 
 import jinja2
 from typing_extensions import TypedDict
 
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.exceptions import UserError
 from daras_ai_v2.scrollable_html_widget import scrollable_html
 
@@ -16,7 +16,7 @@ class SearchReference(TypedDict):
     score: float
 
 
-class CitationStyles(Enum):
+class CitationStyles(GooeyEnum):
     number = "Numbers ( [1] [2] [3] ..)"
     title = "Source Title ( [Source 1] [Source 2] [Source 3] ..)"
     url = "Source URL ( [https://source1.com] [https://source2.com] [https://source3.com] ..)"
diff --git a/daras_ai_v2/serp_search.py b/daras_ai_v2/serp_search.py
index b95586461..7dea257de 100644
--- a/daras_ai_v2/serp_search.py
+++ b/daras_ai_v2/serp_search.py
@@ -20,7 +20,7 @@ def get_related_questions_from_serp_api(
 ) -> tuple[dict, list[str]]:
     data = call_serp_api(
         search_query,
-        search_type=SerpSearchType.SEARCH,
+        search_type=SerpSearchType.search,
         search_location=search_location,
     )
     items = data.get("peopleAlsoAsk", []) or data.get("relatedSearches", [])
@@ -66,10 +66,10 @@ def call_serp_api(
     search_location: SerpSearchLocation,
 ) -> dict:
     r = requests.post(
-        "https://google.serper.dev/" + search_type.value,
+        "https://google.serper.dev/" + search_type.api_value,
         json=dict(
             q=query,
-            gl=search_location.value,
+            gl=search_location.api_value,
         ),
         headers={"X-API-KEY": settings.SERPER_API_KEY},
     )
diff --git a/daras_ai_v2/serp_search_locations.py b/daras_ai_v2/serp_search_locations.py
index 2c26dc8a9..bcb39c053 100644
--- a/daras_ai_v2/serp_search_locations.py
+++ b/daras_ai_v2/serp_search_locations.py
@@ -1,8 +1,11 @@
-from django.db.models import TextChoices
+import typing
+
 from pydantic import BaseModel
 from pydantic import Field
 
 import gooey_gui as gui
+from daras_ai_v2.custom_enum import GooeyEnum
+from daras_ai_v2.enum_selector_widget import enum_selector
 from daras_ai_v2.field_render import field_title_desc
 
 
@@ -26,10 +29,10 @@ def serp_search_settings():
 
 
 def serp_search_type_selectbox(key="serp_search_type"):
-    gui.selectbox(
+    enum_selector(
+        SerpSearchType,
         f"###### {field_title_desc(GoogleSearchMixin, key)}",
-        options=SerpSearchType,
-        format_func=lambda x: x.label,
+        use_selectbox=True,
         key=key,
     )
 
@@ -37,22 +40,31 @@ def serp_search_type_selectbox(key="serp_search_type"):
 def serp_search_location_selectbox(key="serp_search_location"):
     gui.selectbox(
         f"###### {field_title_desc(GoogleSearchMixin, key)}",
-        options=SerpSearchLocation,
-        format_func=lambda x: f"{x.label} ({x.value})",
+        options=[e.api_value for e in SerpSearchLocations],
+        format_func=lambda e: f"{SerpSearchLocations.from_api(e).label} ({e})",
         key=key,
-        value=SerpSearchLocation.UNITED_STATES,
+        value=SerpSearchLocations.UNITED_STATES.api_value,
     )
 
 
-class SerpSearchType(TextChoices):
-    SEARCH = "search", "🔎 Search"
-    IMAGES = "images", "📷 Images"
-    VIDEOS = "videos", "🎥 Videos"
-    PLACES = "places", "📍 Places"
-    NEWS = "news", "📰 News"
+class SerpSearchType(GooeyEnum):
+    search = "🔎 Search"
+    images = "📷 Images"
+    videos = "🎥 Videos"
+    places = "📍 Places"
+    news = "📰 News"
+
+    @property
+    def label(self):
+        return self.value
+
+
+class SerpSearchLocation(typing.NamedTuple):
+    api_value: str
+    label: str
 
 
-class SerpSearchLocation(TextChoices):
+class SerpSearchLocations(SerpSearchLocation, GooeyEnum):
     AFGHANISTAN = "af", "Afghanistan"
     ALBANIA = "al", "Albania"
     ALGERIA = "dz", "Algeria"
@@ -304,7 +316,7 @@ class SerpSearchLocation(TextChoices):
 
 
 class GoogleSearchLocationMixin(BaseModel):
-    serp_search_location: SerpSearchLocation | None = Field(
+    serp_search_location: SerpSearchLocations.api_enum | None = Field(
         title="Web Search Location",
     )
     scaleserp_locations: list[str] | None = Field(
@@ -313,7 +325,7 @@ class GoogleSearchLocationMixin(BaseModel):
 
 
 class GoogleSearchMixin(GoogleSearchLocationMixin, BaseModel):
-    serp_search_type: SerpSearchType | None = Field(
+    serp_search_type: SerpSearchType.api_enum | None = Field(
         title="Web Search Type",
     )
     scaleserp_search_field: str | None = Field(
diff --git a/daras_ai_v2/stable_diffusion.py b/daras_ai_v2/stable_diffusion.py
index f59b044c3..a6ee14422 100644
--- a/daras_ai_v2/stable_diffusion.py
+++ b/daras_ai_v2/stable_diffusion.py
@@ -1,6 +1,5 @@
 import io
 import typing
-from enum import Enum
 
 import requests
 from PIL import Image
@@ -13,6 +12,7 @@
     resize_img_fit,
     get_downscale_factor,
 )
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.exceptions import (
     raise_for_status,
     UserError,
@@ -27,179 +27,249 @@
 SD_IMG_MAX_SIZE = (768, 768)
 
 
-class InpaintingModels(Enum):
-    sd_2 = "Stable Diffusion v2.1 (stability.ai)"
-    runway_ml = "Stable Diffusion v1.5 (RunwayML)"
-    dall_e = "Dall-E (OpenAI)"
+class InpaintingModel(typing.NamedTuple):
+    model_id: str | None
+    label: str
 
-    jack_qiao = "Stable Diffusion v1.4 [Deprecated] (Jack Qiao)"
+
+class InpaintingModels(InpaintingModel, GooeyEnum):
+    sd_2 = InpaintingModel(
+        label="Stable Diffusion v2.1 (stability.ai)",
+        model_id="stabilityai/stable-diffusion-2-inpainting",
+    )
+    runway_ml = InpaintingModel(
+        label="Stable Diffusion v1.5 (RunwayML)",
+        model_id="runwayml/stable-diffusion-inpainting",
+    )
+    dall_e = InpaintingModel(label="Dall-E (OpenAI)", model_id="dall-e-2")
+
+    jack_qiao = InpaintingModel(
+        label="Stable Diffusion v1.4 [Deprecated] (Jack Qiao)", model_id=None
+    )
 
     @classmethod
     def _deprecated(cls):
         return {cls.jack_qiao}
 
 
-inpaint_model_ids = {
-    InpaintingModels.sd_2: "stabilityai/stable-diffusion-2-inpainting",
-    InpaintingModels.runway_ml: "runwayml/stable-diffusion-inpainting",
-}
+class TextToImageModel(typing.NamedTuple):
+    model_id: str | None
+    label: str
 
 
-class Text2ImgModels(Enum):
+class TextToImageModels(TextToImageModel, GooeyEnum):
     # sd_1_4 = "SD v1.4 (RunwayML)" # Host this too?
-    dream_shaper = "DreamShaper (Lykon)"
-    dreamlike_2 = "Dreamlike Photoreal 2.0 (dreamlike.art)"
-    sd_2 = "Stable Diffusion v2.1 (stability.ai)"
-    sd_1_5 = "Stable Diffusion v1.5 (RunwayML)"
+    dream_shaper = TextToImageModel(
+        label="DreamShaper (Lykon)", model_id="Lykon/DreamShaper"
+    )
+    dreamlike_2 = TextToImageModel(
+        label="Dreamlike Photoreal 2.0 (dreamlike.art)",
+        model_id="dreamlike-art/dreamlike-photoreal-2.0",
+    )
+    sd_2 = TextToImageModel(
+        label="Stable Diffusion v2.1 (stability.ai)",
+        model_id="stabilityai/stable-diffusion-2-1",
+    )
+    sd_1_5 = TextToImageModel(
+        label="Stable Diffusion v1.5 (RunwayML)",
+        model_id="runwayml/stable-diffusion-v1-5",
+    )
 
-    dall_e = "DALL·E 2 (OpenAI)"
-    dall_e_3 = "DALL·E 3 (OpenAI)"
+    dall_e = TextToImageModel(label="DALL·E 2 (OpenAI)", model_id="dall-e-2")
+    dall_e_3 = TextToImageModel(label="DALL·E 3 (OpenAI)", model_id="dall-e-3")
 
-    openjourney_2 = "Open Journey v2 beta (PromptHero)"
-    openjourney = "Open Journey (PromptHero)"
-    analog_diffusion = "Analog Diffusion (wavymulder)"
-    protogen_5_3 = "Protogen v5.3 (darkstorm2150)"
+    openjourney_2 = TextToImageModel(
+        label="Open Journey v2 beta (PromptHero)", model_id="prompthero/openjourney-v2"
+    )
+    openjourney = TextToImageModel(
+        label="Open Journey (PromptHero)", model_id="prompthero/openjourney"
+    )
+    analog_diffusion = TextToImageModel(
+        label="Analog Diffusion (wavymulder)", model_id="wavymulder/Analog-Diffusion"
+    )
+    protogen_5_3 = TextToImageModel(
+        label="Protogen v5.3 (darkstorm2150)",
+        model_id="darkstorm2150/Protogen_v5.3_Official_Release",
+    )
 
-    jack_qiao = "Stable Diffusion v1.4 [Deprecated] (Jack Qiao)"
-    rodent_diffusion_1_5 = "Rodent Diffusion 1.5 [Deprecated] (NerdyRodent)"
-    deepfloyd_if = "DeepFloyd IF [Deprecated] (stability.ai)"
+    jack_qiao = TextToImageModel(
+        label="Stable Diffusion v1.4 [Deprecated] (Jack Qiao)", model_id=None
+    )
+    rodent_diffusion_1_5 = TextToImageModel(
+        label="Rodent Diffusion 1.5 [Deprecated] (NerdyRodent)", model_id=None
+    )
+    deepfloyd_if = TextToImageModel(
+        label="DeepFloyd IF [Deprecated] (stability.ai)", model_id=None
+    )
 
     @classmethod
     def _deprecated(cls):
         return {cls.jack_qiao, cls.deepfloyd_if, cls.rodent_diffusion_1_5}
 
 
-text2img_model_ids = {
-    Text2ImgModels.sd_1_5: "runwayml/stable-diffusion-v1-5",
-    Text2ImgModels.sd_2: "stabilityai/stable-diffusion-2-1",
-    Text2ImgModels.dream_shaper: "Lykon/DreamShaper",
-    Text2ImgModels.analog_diffusion: "wavymulder/Analog-Diffusion",
-    Text2ImgModels.openjourney: "prompthero/openjourney",
-    Text2ImgModels.openjourney_2: "prompthero/openjourney-v2",
-    Text2ImgModels.dreamlike_2: "dreamlike-art/dreamlike-photoreal-2.0",
-    Text2ImgModels.protogen_5_3: "darkstorm2150/Protogen_v5.3_Official_Release",
-}
-dall_e_model_ids = {
-    Text2ImgModels.dall_e: "dall-e-2",
-    Text2ImgModels.dall_e_3: "dall-e-3",
-}
+class ImageToImageModel(typing.NamedTuple):
+    model_id: str | None
+    label: str
 
 
-class Img2ImgModels(Enum):
-    dream_shaper = "DreamShaper (Lykon)"
-    dreamlike_2 = "Dreamlike Photoreal 2.0 (dreamlike.art)"
-    sd_2 = "Stable Diffusion v2.1 (stability.ai)"
-    sd_1_5 = "Stable Diffusion v1.5 (RunwayML)"
+class ImageToImageModels(ImageToImageModel, GooeyEnum):
+    dream_shaper = ImageToImageModel(
+        label="DreamShaper (Lykon)", model_id="Lykon/DreamShaper"
+    )
+    dreamlike_2 = ImageToImageModel(
+        label="Dreamlike Photoreal 2.0 (dreamlike.art)",
+        model_id="dreamlike-art/dreamlike-photoreal-2.0",
+    )
+    sd_2 = ImageToImageModel(
+        label="Stable Diffusion v2.1 (stability.ai)",
+        model_id="stabilityai/stable-diffusion-2-1",
+    )
+    sd_1_5 = ImageToImageModel(
+        label="Stable Diffusion v1.5 (RunwayML)",
+        model_id="runwayml/stable-diffusion-v1-5",
+    )
 
-    dall_e = "Dall-E (OpenAI)"
+    dall_e = ImageToImageModel(label="Dall-E (OpenAI)", model_id=None)
 
-    instruct_pix2pix = "✨ InstructPix2Pix (Tim Brooks)"
-    openjourney_2 = "Open Journey v2 beta (PromptHero) 🐢"
-    openjourney = "Open Journey (PromptHero) 🐢"
-    analog_diffusion = "Analog Diffusion (wavymulder) 🐢"
-    protogen_5_3 = "Protogen v5.3 (darkstorm2150) 🐢"
+    instruct_pix2pix = ImageToImageModel(
+        label="✨ InstructPix2Pix (Tim Brooks)", model_id=None
+    )
+    openjourney_2 = ImageToImageModel(
+        label="Open Journey v2 beta (PromptHero) 🐢",
+        model_id="prompthero/openjourney-v2",
+    )
+    openjourney = ImageToImageModel(
+        label="Open Journey (PromptHero) 🐢", model_id="prompthero/openjourney"
+    )
+    analog_diffusion = ImageToImageModel(
+        label="Analog Diffusion (wavymulder) 🐢", model_id="wavymulder/Analog-Diffusion"
+    )
+    protogen_5_3 = ImageToImageModel(
+        label="Protogen v5.3 (darkstorm2150) 🐢",
+        model_id="darkstorm2150/Protogen_v5.3_Official_Release",
+    )
 
-    jack_qiao = "Stable Diffusion v1.4 [Deprecated] (Jack Qiao)"
-    rodent_diffusion_1_5 = "Rodent Diffusion 1.5 [Deprecated] (NerdyRodent)"
+    jack_qiao = ImageToImageModel(
+        label="Stable Diffusion v1.4 [Deprecated] (Jack Qiao)", model_id=None
+    )
+    rodent_diffusion_1_5 = ImageToImageModel(
+        label="Rodent Diffusion 1.5 [Deprecated] (NerdyRodent)", model_id=None
+    )
 
     @classmethod
     def _deprecated(cls):
         return {cls.jack_qiao, cls.rodent_diffusion_1_5}
 
 
-img2img_model_ids = {
-    Img2ImgModels.sd_2: "stabilityai/stable-diffusion-2-1",
-    Img2ImgModels.sd_1_5: "runwayml/stable-diffusion-v1-5",
-    Img2ImgModels.dream_shaper: "Lykon/DreamShaper",
-    Img2ImgModels.openjourney: "prompthero/openjourney",
-    Img2ImgModels.openjourney_2: "prompthero/openjourney-v2",
-    Img2ImgModels.analog_diffusion: "wavymulder/Analog-Diffusion",
-    Img2ImgModels.protogen_5_3: "darkstorm2150/Protogen_v5.3_Official_Release",
-    Img2ImgModels.dreamlike_2: "dreamlike-art/dreamlike-photoreal-2.0",
-}
-
-
-class ControlNetModels(Enum):
-    sd_controlnet_canny = "Canny"
-    sd_controlnet_depth = "Depth"
-    sd_controlnet_hed = "HED Boundary"
-    sd_controlnet_mlsd = "M-LSD Straight Line"
-    sd_controlnet_normal = "Normal Map"
-    sd_controlnet_openpose = "Human Pose"
-    sd_controlnet_scribble = "Scribble"
-    sd_controlnet_seg = "Image Segmentation"
-    sd_controlnet_tile = "Tiling"
-    sd_controlnet_brightness = "Brightness"
-    control_v1p_sd15_qrcode_monster_v2 = "QR Monster V2"
-
-
-controlnet_model_explanations = {
-    ControlNetModels.sd_controlnet_canny: "Canny edge detection",
-    ControlNetModels.sd_controlnet_depth: "Depth estimation",
-    ControlNetModels.sd_controlnet_hed: "HED edge detection",
-    ControlNetModels.sd_controlnet_mlsd: "M-LSD straight line detection",
-    ControlNetModels.sd_controlnet_normal: "Normal map estimation",
-    ControlNetModels.sd_controlnet_openpose: "Human pose estimation",
-    ControlNetModels.sd_controlnet_scribble: "Scribble",
-    ControlNetModels.sd_controlnet_seg: "Image segmentation",
-    ControlNetModels.sd_controlnet_tile: "Tiling: to preserve small details",
-    ControlNetModels.sd_controlnet_brightness: "Brightness: to increase contrast naturally",
-    ControlNetModels.control_v1p_sd15_qrcode_monster_v2: "QR Monster: make beautiful QR codes that still scan with a controlnet specifically trained for this purpose",
-}
-
-controlnet_model_ids = {
-    ControlNetModels.sd_controlnet_canny: "lllyasviel/sd-controlnet-canny",
-    ControlNetModels.sd_controlnet_depth: "lllyasviel/sd-controlnet-depth",
-    ControlNetModels.sd_controlnet_hed: "lllyasviel/sd-controlnet-hed",
-    ControlNetModels.sd_controlnet_mlsd: "lllyasviel/sd-controlnet-mlsd",
-    ControlNetModels.sd_controlnet_normal: "lllyasviel/sd-controlnet-normal",
-    ControlNetModels.sd_controlnet_openpose: "lllyasviel/sd-controlnet-openpose",
-    ControlNetModels.sd_controlnet_scribble: "lllyasviel/sd-controlnet-scribble",
-    ControlNetModels.sd_controlnet_seg: "lllyasviel/sd-controlnet-seg",
-    ControlNetModels.sd_controlnet_tile: "lllyasviel/control_v11f1e_sd15_tile",
-    ControlNetModels.sd_controlnet_brightness: "ioclab/control_v1p_sd15_brightness",
-    ControlNetModels.control_v1p_sd15_qrcode_monster_v2: "monster-labs/control_v1p_sd15_qrcode_monster/v2",
-}
-
-
-class Schedulers(models.TextChoices):
-    singlestep_dpm_solver = (
-        "DPM",
-        "DPMSolverSinglestepScheduler",
-    )
-    multistep_dpm_solver = "DPM Multistep", "DPMSolverMultistepScheduler"
-    dpm_sde = (
-        "DPM SDE",
-        "DPMSolverSDEScheduler",
-    )
-    dpm_discrete = (
-        "DPM Discrete",
-        "KDPM2DiscreteScheduler",
-    )
-    dpm_discrete_ancestral = (
-        "DPM Anscetral",
-        "KDPM2AncestralDiscreteScheduler",
-    )
-    unipc = "UniPC", "UniPCMultistepScheduler"
-    lms_discrete = (
-        "LMS",
-        "LMSDiscreteScheduler",
-    )
-    heun = (
-        "Heun",
-        "HeunDiscreteScheduler",
-    )
-    euler = "Euler", "EulerDiscreteScheduler"
-    euler_ancestral = (
-        "Euler ancestral",
-        "EulerAncestralDiscreteScheduler",
-    )
-    pndm = "PNDM", "PNDMScheduler"
-    ddpm = "DDPM", "DDPMScheduler"
-    ddim = "DDIM", "DDIMScheduler"
-    deis = (
-        "DEIS",
-        "DEISMultistepScheduler",
+class ControlNetModel(typing.NamedTuple):
+    label: str
+    model_id: str
+    explanation: str
+
+
+class ControlNetModels(ControlNetModel, GooeyEnum):
+    sd_controlnet_canny = ControlNetModel(
+        label="Canny",
+        explanation="Canny edge detection",
+        model_id="lllyasviel/sd-controlnet-canny",
+    )
+    sd_controlnet_depth = ControlNetModel(
+        label="Depth",
+        explanation="Depth estimation",
+        model_id="lllyasviel/sd-controlnet-depth",
+    )
+    sd_controlnet_hed = ControlNetModel(
+        label="HED Boundary",
+        explanation="HED edge detection",
+        model_id="lllyasviel/sd-controlnet-hed",
+    )
+    sd_controlnet_mlsd = ControlNetModel(
+        label="M-LSD Straight Line",
+        explanation="M-LSD straight line detection",
+        model_id="lllyasviel/sd-controlnet-mlsd",
+    )
+    sd_controlnet_normal = ControlNetModel(
+        label="Normal Map",
+        explanation="Normal map estimation",
+        model_id="lllyasviel/sd-controlnet-normal",
+    )
+    sd_controlnet_openpose = ControlNetModel(
+        label="Human Pose",
+        explanation="Human pose estimation",
+        model_id="lllyasviel/sd-controlnet-openpose",
+    )
+    sd_controlnet_scribble = ControlNetModel(
+        label="Scribble",
+        explanation="Scribble",
+        model_id="lllyasviel/sd-controlnet-scribble",
+    )
+    sd_controlnet_seg = ControlNetModel(
+        label="Image Segmentation",
+        explanation="Image segmentation",
+        model_id="lllyasviel/sd-controlnet-seg",
+    )
+    sd_controlnet_tile = ControlNetModel(
+        label="Tiling",
+        explanation="Tiling: to preserve small details",
+        model_id="lllyasviel/control_v11f1e_sd15_tile",
+    )
+    sd_controlnet_brightness = ControlNetModel(
+        label="Brightness",
+        explanation="Brightness: to increase contrast naturally",
+        model_id="ioclab/control_v1p_sd15_brightness",
+    )
+    control_v1p_sd15_qrcode_monster_v2 = ControlNetModel(
+        label="QR Monster V2",
+        explanation="QR Monster: make beautiful QR codes that still scan with a controlnet specifically trained for this purpose",
+        model_id="monster-labs/control_v1p_sd15_qrcode_monster/v2",
+    )
+
+
+class Scheduler(typing.NamedTuple):
+    label: str
+    model_id: str
+
+
+class Schedulers(Scheduler, GooeyEnum):
+    singlestep_dpm_solver = Scheduler(
+        label="DPM",
+        model_id="DPMSolverSinglestepScheduler",
+    )
+    multistep_dpm_solver = Scheduler(
+        label="DPM Multistep", model_id="DPMSolverMultistepScheduler"
+    )
+    dpm_sde = Scheduler(
+        label="DPM SDE",
+        model_id="DPMSolverSDEScheduler",
+    )
+    dpm_discrete = Scheduler(
+        label="DPM Discrete",
+        model_id="KDPM2DiscreteScheduler",
+    )
+    dpm_discrete_ancestral = Scheduler(
+        label="DPM Anscetral",
+        model_id="KDPM2AncestralDiscreteScheduler",
+    )
+    unipc = Scheduler(label="UniPC", model_id="UniPCMultistepScheduler")
+    lms_discrete = Scheduler(
+        label="LMS",
+        model_id="LMSDiscreteScheduler",
+    )
+    heun = Scheduler(
+        label="Heun",
+        model_id="HeunDiscreteScheduler",
+    )
+    euler = Scheduler("Euler", "EulerDiscreteScheduler")
+    euler_ancestral = Scheduler(
+        label="Euler ancestral",
+        model_id="EulerAncestralDiscreteScheduler",
+    )
+    pndm = Scheduler(label="PNDM", model_id="PNDMScheduler")
+    ddpm = Scheduler(label="DDPM", model_id="DDPMScheduler")
+    ddim = Scheduler(label="DDIM", model_id="DDIMScheduler")
+    deis = Scheduler(
+        label="DEIS",
+        model_id="DEISMultistepScheduler",
     )
 
 
@@ -282,18 +352,18 @@ def text2img(
     dall_e_3_quality: str | None = None,
     dall_e_3_style: str | None = None,
 ):
-    if selected_model != Text2ImgModels.dall_e_3.name:
+    if selected_model != TextToImageModels.dall_e_3.name:
         _resolution_check(width, height, max_size=(1024, 1024))
 
     match selected_model:
-        case Text2ImgModels.dall_e_3.name:
+        case TextToImageModels.dall_e_3.name:
             from openai import OpenAI
 
             client = OpenAI()
             width, height = _get_dall_e_3_img_size(width, height)
             with capture_openai_content_policy_violation():
                 response = client.images.generate(
-                    model=dall_e_model_ids[Text2ImgModels[selected_model]],
+                    model=TextToImageModels[selected_model].model_id,
                     n=1,  # num_outputs, not supported yet
                     prompt=prompt,
                     response_format="b64_json",
@@ -302,7 +372,7 @@ def text2img(
                     size=f"{width}x{height}",
                 )
             out_imgs = [b64_img_decode(part.b64_json) for part in response.data]
-        case Text2ImgModels.dall_e.name:
+        case TextToImageModels.dall_e.name:
             from openai import OpenAI
 
             edge = _get_dall_e_img_size(width, height)
@@ -320,8 +390,8 @@ def text2img(
             return call_sd_multi(
                 "diffusion.text2img",
                 pipeline={
-                    "model_id": text2img_model_ids[Text2ImgModels[selected_model]],
-                    "scheduler": Schedulers[scheduler].label if scheduler else None,
+                    "model_id": TextToImageModels[selected_model].model_id,
+                    "scheduler": Schedulers[scheduler].model_id if scheduler else None,
                     "disable_safety_checker": True,
                     "seed": seed,
                 },
@@ -382,7 +452,7 @@ def img2img(
     _resolution_check(width, height)
 
     match selected_model:
-        case Img2ImgModels.dall_e.name:
+        case ImageToImageModels.dall_e.name:
             from openai import OpenAI
 
             edge = _get_dall_e_img_size(width, height)
@@ -413,7 +483,7 @@ def img2img(
             return call_sd_multi(
                 "diffusion.img2img",
                 pipeline={
-                    "model_id": img2img_model_ids[Img2ImgModels[selected_model]],
+                    "model_id": ImageToImageModels[selected_model].model_id,
                     # "scheduler": "UniPCMultistepScheduler",
                     "disable_safety_checker": True,
                     "seed": seed,
@@ -456,15 +526,16 @@ def controlnet(
     return call_sd_multi(
         "diffusion.controlnet",
         pipeline={
-            "model_id": text2img_model_ids[Text2ImgModels[selected_model]],
+            "model_id": TextToImageModels[selected_model].model_id,
             "seed": seed,
             "scheduler": (
-                Schedulers[scheduler].label if scheduler else "UniPCMultistepScheduler"
+                Schedulers[scheduler].model_id
+                if scheduler
+                else Schedulers.unipc.model_id
             ),
             "disable_safety_checker": True,
             "controlnet_model_id": [
-                controlnet_model_ids[ControlNetModels[model]]
-                for model in selected_controlnet_model
+                ControlNetModels[model].model_id for model in selected_controlnet_model
             ],
         },
         inputs={
@@ -482,13 +553,13 @@ def controlnet(
 
 def add_prompt_prefix(prompt: str, selected_model: str) -> str:
     match selected_model:
-        case Text2ImgModels.openjourney.name:
+        case TextToImageModels.openjourney.name:
             prompt = "mdjrny-v4 style " + prompt
-        case Text2ImgModels.analog_diffusion.name:
+        case TextToImageModels.analog_diffusion.name:
             prompt = "analog style " + prompt
-        case Text2ImgModels.protogen_5_3.name:
+        case TextToImageModels.protogen_5_3.name:
             prompt = "modelshoot style " + prompt
-        case Text2ImgModels.dreamlike_2.name:
+        case TextToImageModels.dreamlike_2.name:
             prompt = "photo, " + prompt
     return prompt
 
@@ -535,9 +606,9 @@ def inpainting(
             out_imgs_urls = call_sd_multi(
                 "diffusion.inpaint",
                 pipeline={
-                    "model_id": inpaint_model_ids[InpaintingModels[selected_model]],
+                    "model_id": InpaintingModels[selected_model].model_id,
                     "seed": seed,
-                    # "scheduler": Schedulers[scheduler].label
+                    # "scheduler": Schedulers[scheduler].model_id
                     # if scheduler
                     # else "UniPCMultistepScheduler",
                     "disable_safety_checker": True,
diff --git a/daras_ai_v2/text_to_speech_settings_widgets.py b/daras_ai_v2/text_to_speech_settings_widgets.py
index 5214d5cc5..65b5b56ea 100644
--- a/daras_ai_v2/text_to_speech_settings_widgets.py
+++ b/daras_ai_v2/text_to_speech_settings_widgets.py
@@ -46,7 +46,7 @@ class OpenAI_TTS_Voices(GooeyEnum):
     shimmer = "shimmer"
 
 
-class TextToSpeechProviders(Enum):
+class TextToSpeechProviders(GooeyEnum):
     GOOGLE_TTS = "Google Text-to-Speech"
     ELEVEN_LABS = "Eleven Labs"
     UBERDUCK = "Uberduck.ai"
diff --git a/daras_ai_v2/upscaler_models.py b/daras_ai_v2/upscaler_models.py
index 11d8ff225..f5d46a40a 100644
--- a/daras_ai_v2/upscaler_models.py
+++ b/daras_ai_v2/upscaler_models.py
@@ -1,11 +1,11 @@
 import typing
-from enum import Enum
 from pathlib import Path
 
 import replicate
 import requests
 
 from daras_ai.image_input import upload_file_from_bytes
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.exceptions import UserError
 from daras_ai_v2.gpu_server import call_celery_task_outfile
 from daras_ai_v2.pydantic_validation import FieldHttpUrl
@@ -19,7 +19,7 @@ class UpscalerModel(typing.NamedTuple):
     is_bg_model: bool = False
 
 
-class UpscalerModels(UpscalerModel, Enum):
+class UpscalerModels(UpscalerModel, GooeyEnum):
     gfpgan_1_4 = UpscalerModel(
         model_id="GFPGANv1.4",
         label="GFPGAN v1.4 (Tencent ARC)",
diff --git a/daras_ai_v2/vector_search.py b/daras_ai_v2/vector_search.py
index 36de65f29..ba0c4ce59 100644
--- a/daras_ai_v2/vector_search.py
+++ b/daras_ai_v2/vector_search.py
@@ -71,8 +71,13 @@
 
 
 class DocSearchRequest(BaseModel):
+    class Config:
+        use_enum_values = True
+
     search_query: str
-    keyword_query: str | list[str] | None
+    keyword_query: str | list[str] | None = Field(
+        **{"x-fern-type-name": "KeywordQuery"}
+    )
 
     documents: list[str] | None
 
@@ -82,7 +87,7 @@ class DocSearchRequest(BaseModel):
 
     doc_extract_url: str | None
 
-    embedding_model: typing.Literal[tuple(e.name for e in EmbeddingModels)] | None
+    embedding_model: EmbeddingModels.api_enum | None
     dense_weight: float | None = Field(
         ge=0.0,
         le=1.0,
diff --git a/embeddings/models.py b/embeddings/models.py
index f30b1a0bf..f5f350bbc 100644
--- a/embeddings/models.py
+++ b/embeddings/models.py
@@ -28,7 +28,7 @@ class EmbeddedFile(models.Model):
     selected_asr_model = models.CharField(max_length=100, blank=True)
     embedding_model = models.CharField(
         max_length=100,
-        choices=[(model.name, model.label) for model in EmbeddingModels],
+        choices=EmbeddingModels.db_choices(),
         default=EmbeddingModels.openai_3_large.name,
     )
 
diff --git a/functions/models.py b/functions/models.py
index be1be7a5e..711805d64 100644
--- a/functions/models.py
+++ b/functions/models.py
@@ -22,7 +22,7 @@ class RecipeFunction(BaseModel):
         title="URL",
         description="The URL of the [function](https://gooey.ai/functions) to call.",
     )
-    trigger: FunctionTrigger.api_choices = Field(
+    trigger: FunctionTrigger.api_enum = Field(
         title="Trigger",
         description="When to run this function. `pre` runs before the recipe, `post` runs after the recipe.",
     )
@@ -30,7 +30,7 @@ class RecipeFunction(BaseModel):
 
 class CalledFunctionResponse(BaseModel):
     url: str
-    trigger: FunctionTrigger.api_choices
+    trigger: FunctionTrigger.api_enum
     return_value: typing.Any
 
     @classmethod
diff --git a/recipes/BulkEval.py b/recipes/BulkEval.py
index 7f92a53e1..62aabe8c3 100644
--- a/recipes/BulkEval.py
+++ b/recipes/BulkEval.py
@@ -139,6 +139,7 @@ class BulkEvalPage(BasePage):
     title = "Evaluator"
     workflow = Workflow.BULK_EVAL
     slug_versions = ["bulk-eval", "eval"]
+    sdk_method_name = "eval"
 
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/aad314f0-9a97-11ee-8318-02420a0001c7/W.I.9.png.png"
 
@@ -185,9 +186,7 @@ class RequestModelBase(BasePage.RequestModel):
             """,
         )
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
     class RequestModel(LanguageModelSettings, RequestModelBase):
         pass
diff --git a/recipes/BulkRunner.py b/recipes/BulkRunner.py
index 7ad9e67e9..7ab4e4358 100644
--- a/recipes/BulkRunner.py
+++ b/recipes/BulkRunner.py
@@ -43,6 +43,8 @@ class BulkRunnerPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/87f35df4-88d7-11ee-aac9-02420a00016b/Bulk%20Runner.png.png"
     workflow = Workflow.BULK_RUNNER
     slug_versions = ["bulk-runner", "bulk"]
+    sdk_method_name = "bulkRun"
+
     price = 1
 
     class RequestModel(BasePage.RequestModel):
diff --git a/recipes/ChyronPlant.py b/recipes/ChyronPlant.py
index fef4714b9..0777b93c7 100644
--- a/recipes/ChyronPlant.py
+++ b/recipes/ChyronPlant.py
@@ -1,4 +1,5 @@
 import gooey_gui as gui
+from gooey_gui.components import typing
 from pydantic import BaseModel
 
 from bots.models import Workflow
@@ -12,6 +13,7 @@ class ChyronPlantPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/aeb83ee8-889e-11ee-93dc-02420a000143/Youtube%20transcripts%20GPT%20extractions.png.png"
     workflow = Workflow.CHYRON_PLANT
     slug_versions = ["ChyronPlant"]
+    sdk_method_name = ""
 
     class RequestModel(BasePage.RequestModel):
         midi_notes: str
@@ -30,6 +32,10 @@ class ResponseModel(BaseModel):
         midi_translation: str
         chyron_output: str
 
+    @classmethod
+    def get_openapi_extra(cls) -> dict[str, typing.Any]:
+        return {"x-fern-ignore": True}
+
     def render_form_v2(self):
         gui.text_input(
             """
diff --git a/recipes/CompareLLM.py b/recipes/CompareLLM.py
index 513421146..55d317176 100644
--- a/recipes/CompareLLM.py
+++ b/recipes/CompareLLM.py
@@ -30,6 +30,7 @@ class CompareLLMPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/ae42015e-88d7-11ee-aac9-02420a00016b/Compare%20LLMs.png.png"
     workflow = Workflow.COMPARE_LLM
     slug_versions = ["CompareLLM", "llm", "compare-large-language-models"]
+    sdk_method_name = "llm"
 
     functions_in_settings = False
 
@@ -43,17 +44,16 @@ class CompareLLMPage(BasePage):
 
     class RequestModelBase(BasePage.RequestModel):
         input_prompt: str | None
-        selected_models: (
-            list[typing.Literal[tuple(e.name for e in LargeLanguageModels)]] | None
-        )
+        selected_models: list[LargeLanguageModels.api_enum] | None
 
     class RequestModel(LanguageModelSettings, RequestModelBase):
         pass
 
     class ResponseModel(BaseModel):
-        output_text: dict[
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)], list[str]
-        ]
+        class Config:
+            use_enum_values = True
+
+        output_text: dict[LargeLanguageModels.api_enum, list[str]]
 
     def preview_image(self, state: dict) -> str | None:
         return DEFAULT_COMPARE_LM_META_IMG
diff --git a/recipes/CompareText2Img.py b/recipes/CompareText2Img.py
index f41a46170..ee0e985fb 100644
--- a/recipes/CompareText2Img.py
+++ b/recipes/CompareText2Img.py
@@ -20,7 +20,7 @@
 from daras_ai_v2.loom_video_widget import youtube_video
 from daras_ai_v2.safety_checker import safety_checker
 from daras_ai_v2.stable_diffusion import (
-    Text2ImgModels,
+    TextToImageModels,
     text2img,
     instruct_pix2pix,
     sd_upscale,
@@ -39,6 +39,7 @@ class CompareText2ImgPage(BasePage):
         "text2img",
         "compare-ai-image-generators",
     ]
+    sdk_method_name = "textToImage"
 
     sane_defaults = {
         "guidance_scale": 7.5,
@@ -65,19 +66,14 @@ class RequestModel(BasePage.RequestModel):
         seed: int | None
         sd_2_upscaling: bool | None
 
-        selected_models: (
-            list[typing.Literal[tuple(e.name for e in Text2ImgModels)]] | None
-        )
-        scheduler: typing.Literal[tuple(e.name for e in Schedulers)] | None
+        selected_models: list[TextToImageModels.api_enum] | None
+        scheduler: Schedulers.api_enum | None
 
         edit_instruction: str | None
         image_guidance_scale: float | None
 
     class ResponseModel(BaseModel):
-        output_images: dict[
-            typing.Literal[tuple(e.name for e in Text2ImgModels)],
-            list[FieldHttpUrl],
-        ]
+        output_images: dict[TextToImageModels.api_enum, list[FieldHttpUrl]]
 
     @classmethod
     def get_example_preferred_fields(cls, state: dict) -> list[str]:
@@ -124,7 +120,7 @@ def render_form_v2(self):
         """
         )
         enum_multiselect(
-            Text2ImgModels,
+            TextToImageModels,
             key="selected_models",
         )
 
@@ -192,7 +188,7 @@ def run(self, state: dict) -> typing.Iterator[str | None]:
         state["output_images"] = output_images = {}
 
         for selected_model in request.selected_models:
-            yield f"Running {Text2ImgModels[selected_model].value}..."
+            yield f"Running {TextToImageModels[selected_model].label}..."
 
             output_images[selected_model] = text2img(
                 selected_model=selected_model,
@@ -253,7 +249,7 @@ def _render_outputs(self, state):
             output_images: dict = state.get("output_images", {}).get(key, [])
             for img in output_images:
                 gui.image(
-                    img, caption=Text2ImgModels[key].value, show_download_button=True
+                    img, caption=TextToImageModels[key].label, show_download_button=True
                 )
 
     def preview_description(self, state: dict) -> str:
@@ -264,9 +260,9 @@ def get_raw_price(self, state: dict) -> int:
         total = 0
         for name in selected_models:
             match name:
-                case Text2ImgModels.deepfloyd_if.name:
+                case TextToImageModels.deepfloyd_if.name:
                     total += 5
-                case Text2ImgModels.dall_e.name | Text2ImgModels.dall_e_3.name:
+                case TextToImageModels.dall_e.name | TextToImageModels.dall_e_3.name:
                     total += 15
                 case _:
                     total += 2
diff --git a/recipes/CompareUpscaler.py b/recipes/CompareUpscaler.py
index 4684ab309..e7032095c 100644
--- a/recipes/CompareUpscaler.py
+++ b/recipes/CompareUpscaler.py
@@ -20,6 +20,7 @@ class CompareUpscalerPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/64393e0c-88db-11ee-b428-02420a000168/AI%20Image%20Upscaler.png.png"
     workflow = Workflow.COMPARE_UPSCALER
     slug_versions = ["compare-ai-upscalers"]
+    sdk_method_name = "upscale"
 
     class RequestModel(BasePage.RequestModel):
         input_image: FieldHttpUrl | None = Field(None, description="Input Image")
@@ -29,21 +30,22 @@ class RequestModel(BasePage.RequestModel):
             description="The final upsampling scale of the image", ge=1, le=4
         )
 
-        selected_models: (
-            list[typing.Literal[tuple(e.name for e in UpscalerModels)]] | None
-        )
+        selected_models: list[UpscalerModels.api_enum] | None
         selected_bg_model: (
             typing.Literal[tuple(e.name for e in UpscalerModels if e.is_bg_model)]
             | None
+        ) = Field(
+            title="Selected Background Model",
+            **{"x-fern-type-name": "BackgroundUpscalerModels"},
         )
 
     class ResponseModel(BaseModel):
-        output_images: dict[
-            typing.Literal[tuple(e.name for e in UpscalerModels)], FieldHttpUrl
-        ] = Field({}, description="Output Images")
-        output_videos: dict[
-            typing.Literal[tuple(e.name for e in UpscalerModels)], FieldHttpUrl
-        ] = Field({}, description="Output Videos")
+        output_images: dict[UpscalerModels.api_enum, FieldHttpUrl] = Field(
+            default_factory=dict, description="Output Images"
+        )
+        output_videos: dict[UpscalerModels.api_enum, FieldHttpUrl] = Field(
+            default_factory=dict, description="Output Videos"
+        )
 
     def validate_form_v2(self):
         assert gui.session_state.get(
@@ -69,6 +71,7 @@ def run_v2(
         for selected_model in request.selected_models:
             model = UpscalerModels[selected_model]
             yield f"Running {model.label}..."
+            print(f"{request.input_image=}, {request.input_video=}")
             if request.input_image:
                 response.output_images[selected_model] = run_upscaler_model(
                     selected_model=model,
diff --git a/recipes/DeforumSD.py b/recipes/DeforumSD.py
index ef4b186b5..eba951bf4 100644
--- a/recipes/DeforumSD.py
+++ b/recipes/DeforumSD.py
@@ -1,8 +1,8 @@
 import typing
 import uuid
 
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.pydantic_validation import FieldHttpUrl
-from django.db.models import TextChoices
 from pydantic import BaseModel
 from typing_extensions import TypedDict
 
@@ -18,17 +18,28 @@
 DEFAULT_DEFORUMSD_META_IMG = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/7dc25196-93fe-11ee-9e3a-02420a0001ce/AI%20Animation%20generator.jpg.png"
 
 
-class AnimationModels(TextChoices):
-    protogen_2_2 = ("Protogen_V2.2.ckpt", "Protogen V2.2 (darkstorm2150)")
-    epicdream = ("epicdream.safetensors", "epiCDream (epinikion)")
+class AnimationModel(typing.NamedTuple):
+    model_id: str
+    label: str
 
 
-class _AnimationPrompt(TypedDict):
+class AnimationModels(AnimationModel, GooeyEnum):
+    protogen_2_2 = AnimationModel(
+        model_id="Protogen_V2.2.ckpt",
+        label="Protogen V2.2 (darkstorm2150)",
+    )
+    epicdream = AnimationModel(
+        model_id="epicdream.safetensors",
+        label="epiCDream (epinikion)",
+    )
+
+
+class AnimationPrompt(TypedDict):
     frame: str
     prompt: str
 
 
-AnimationPrompts = list[_AnimationPrompt]
+AnimationPrompts = list[AnimationPrompt]
 
 CREDITS_PER_FRAME = 1.5
 MODEL_ESTIMATED_TIME_PER_FRAME = 2.4  # seconds
@@ -166,6 +177,7 @@ class DeforumSDPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/media/users/kxmNIYAOJbfOURxHBKNCWeUSKiP2/dd88c110-88d6-11ee-9b4f-2b58bd50e819/animation.gif"
     workflow = Workflow.DEFORUM_SD
     slug_versions = ["DeforumSD", "animation-generator"]
+    sdk_method_name = "animate"
 
     sane_defaults = dict(
         zoom="0: (1.004)",
@@ -185,7 +197,7 @@ class RequestModel(BasePage.RequestModel):
         animation_prompts: AnimationPrompts
         max_frames: int | None
 
-        selected_model: typing.Literal[tuple(e.name for e in AnimationModels)] | None
+        selected_model: AnimationModels.api_enum | None
 
         animation_mode: str | None
         zoom: str | None
@@ -459,7 +471,7 @@ def run(self, state: dict):
             state["output_video"] = call_celery_task_outfile(
                 "deforum",
                 pipeline=dict(
-                    model_id=AnimationModels[request.selected_model].value,
+                    model_id=AnimationModels[request.selected_model].model_id,
                     seed=request.seed,
                 ),
                 inputs=dict(
diff --git a/recipes/DocExtract.py b/recipes/DocExtract.py
index 323e3eab9..9a41f4f19 100644
--- a/recipes/DocExtract.py
+++ b/recipes/DocExtract.py
@@ -86,6 +86,7 @@ class DocExtractPage(BasePage):
         "youtube-bot",
         "doc-extract",
     ]
+    sdk_method_name = "synthesizeData"
     price = 500
 
     class RequestModelBase(BasePage.RequestModel):
@@ -93,7 +94,7 @@ class RequestModelBase(BasePage.RequestModel):
 
         sheet_url: FieldHttpUrl | None
 
-        selected_asr_model: typing.Literal[tuple(e.name for e in AsrModels)] | None
+        selected_asr_model: AsrModels.api_enum | None
         # language: str | None
         google_translate_target: str | None
         glossary_document: FieldHttpUrl | None = Field(
@@ -104,9 +105,7 @@ class RequestModelBase(BasePage.RequestModel):
 
         task_instructions: str | None
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
     class RequestModel(LanguageModelSettings, RequestModelBase):
         pass
diff --git a/recipes/DocSearch.py b/recipes/DocSearch.py
index d0c49f3c7..140ea796e 100644
--- a/recipes/DocSearch.py
+++ b/recipes/DocSearch.py
@@ -2,7 +2,7 @@
 import typing
 
 from furl import furl
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
 import gooey_gui as gui
 from bots.models import Workflow
@@ -52,6 +52,7 @@ class DocSearchPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/cbbb4dc6-88d7-11ee-bf6c-02420a000166/Search%20your%20docs%20with%20gpt.png.png"
     workflow = Workflow.DOC_SEARCH
     slug_versions = ["doc-search"]
+    sdk_method_name = "rag"
 
     sane_defaults = {
         "sampling_temperature": 0.1,
@@ -71,11 +72,9 @@ class RequestModelBase(DocSearchRequest, BasePage.RequestModel):
         task_instructions: str | None
         query_instructions: str | None
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
-        citation_style: typing.Literal[tuple(e.name for e in CitationStyles)] | None
+        citation_style: CitationStyles.api_enum | None
 
     class RequestModel(LanguageModelSettings, RequestModelBase):
         pass
diff --git a/recipes/DocSummary.py b/recipes/DocSummary.py
index 18412e197..1711de053 100644
--- a/recipes/DocSummary.py
+++ b/recipes/DocSummary.py
@@ -1,8 +1,8 @@
 import typing
-from enum import Enum
 
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.pydantic_validation import FieldHttpUrl
-from pydantic import BaseModel, Field
+from pydantic import BaseModel
 
 import gooey_gui as gui
 from bots.models import Workflow
@@ -16,7 +16,6 @@
     LargeLanguageModels,
     run_language_model,
     calc_gpt_tokens,
-    ResponseFormatType,
 )
 from daras_ai_v2.language_model_settings_widgets import (
     language_model_settings,
@@ -38,7 +37,7 @@
 DEFAULT_DOC_SUMMARY_META_IMG = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/f35796d2-93fe-11ee-b86c-02420a0001c7/Summarize%20with%20GPT.jpg.png"
 
 
-class CombineDocumentsChains(Enum):
+class CombineDocumentsChains(GooeyEnum):
     map_reduce = "Map Reduce"
     # refine = "Refine"
     # stuff = "Stuffing (Only works for small documents)"
@@ -49,6 +48,7 @@ class DocSummaryPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/1f858a7a-88d8-11ee-a658-02420a000163/Summarize%20your%20docs%20with%20gpt.png.png"
     workflow = Workflow.DOC_SUMMARY
     slug_versions = ["doc-summary"]
+    sdk_method_name = "docSummary"
 
     price = 225
 
@@ -68,13 +68,11 @@ class RequestModelBase(BasePage.RequestModel):
         task_instructions: str | None
         merge_instructions: str | None
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
-        chain_type: typing.Literal[tuple(e.name for e in CombineDocumentsChains)] | None
+        chain_type: CombineDocumentsChains.api_enum | None
 
-        selected_asr_model: typing.Literal[tuple(e.name for e in AsrModels)] | None
+        selected_asr_model: AsrModels.api_enum | None
         google_translate_target: str | None
 
     class RequestModel(LanguageModelSettings, RequestModelBase):
diff --git a/recipes/EmailFaceInpainting.py b/recipes/EmailFaceInpainting.py
index 3ebc161a9..0fcf39835 100644
--- a/recipes/EmailFaceInpainting.py
+++ b/recipes/EmailFaceInpainting.py
@@ -27,6 +27,7 @@ class EmailFaceInpaintingPage(FaceInpaintingPage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/ec0df5aa-9521-11ee-93d3-02420a0001e5/Email%20Profile%20Lookup.png.png"
     workflow = Workflow.EMAIL_FACE_INPAINTING
     slug_versions = ["EmailFaceInpainting", "ai-image-from-email-lookup"]
+    sdk_method_name = "imageFromEmail"
 
     sane_defaults = {
         "num_outputs": 1,
@@ -49,7 +50,7 @@ class RequestModel(BasePage.RequestModel):
         face_pos_x: float | None
         face_pos_y: float | None
 
-        selected_model: typing.Literal[tuple(e.name for e in InpaintingModels)] | None
+        selected_model: InpaintingModels.api_enum | None
 
         negative_prompt: str | None
 
diff --git a/recipes/FaceInpainting.py b/recipes/FaceInpainting.py
index 8770740fb..a28041d37 100644
--- a/recipes/FaceInpainting.py
+++ b/recipes/FaceInpainting.py
@@ -33,6 +33,7 @@ class FaceInpaintingPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/10c2ce06-88da-11ee-b428-02420a000168/ai%20image%20with%20a%20face.png.png"
     workflow = Workflow.FACE_INPAINTING
     slug_versions = ["FaceInpainting", "face-in-ai-generated-photo"]
+    sdk_method_name = "portrait"
 
     sane_defaults = {
         "num_outputs": 1,
@@ -52,7 +53,7 @@ class RequestModel(BasePage.RequestModel):
         face_pos_x: float | None
         face_pos_y: float | None
 
-        selected_model: typing.Literal[tuple(e.name for e in InpaintingModels)] | None
+        selected_model: InpaintingModels.api_enum | None
 
         negative_prompt: str | None
 
diff --git a/recipes/Functions.py b/recipes/Functions.py
index 81b99c946..434effd27 100644
--- a/recipes/Functions.py
+++ b/recipes/Functions.py
@@ -22,6 +22,7 @@ class FunctionsPage(BasePage):
     title = "Functions"
     workflow = Workflow.FUNCTIONS
     slug_versions = ["functions", "tools", "function", "fn", "functions"]
+    sdk_method_name = "functions"
     show_settings = False
     price = 1
 
diff --git a/recipes/GoogleGPT.py b/recipes/GoogleGPT.py
index 611483287..93d7a0ed2 100644
--- a/recipes/GoogleGPT.py
+++ b/recipes/GoogleGPT.py
@@ -1,7 +1,7 @@
 import typing
 
 from furl import furl
-from pydantic import BaseModel, Field
+from pydantic import BaseModel
 
 import gooey_gui as gui
 from bots.models import Workflow
@@ -11,11 +11,7 @@
     doc_search_advanced_settings,
 )
 from daras_ai_v2.embedding_model import EmbeddingModels
-from daras_ai_v2.language_model import (
-    run_language_model,
-    LargeLanguageModels,
-    ResponseFormatType,
-)
+from daras_ai_v2.language_model import run_language_model, LargeLanguageModels
 from daras_ai_v2.language_model_settings_widgets import (
     language_model_settings,
     language_model_selector,
@@ -31,9 +27,9 @@
 from daras_ai_v2.serp_search import get_links_from_serp_api
 from daras_ai_v2.serp_search_locations import (
     GoogleSearchMixin,
-    serp_search_settings,
-    SerpSearchLocation,
+    SerpSearchLocations,
     SerpSearchType,
+    serp_search_settings,
 )
 from daras_ai_v2.vector_search import render_sources_widget
 from recipes.DocSearch import (
@@ -51,6 +47,7 @@ class GoogleGPTPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/28649544-9406-11ee-bba3-02420a0001cc/Websearch%20GPT%20option%202.png.png"
     workflow = Workflow.GOOGLE_GPT
     slug_versions = ["google-gpt"]
+    sdk_method_name = "webSearchLLM"
 
     price = 175
 
@@ -59,8 +56,8 @@ class GoogleGPTPage(BasePage):
         keywords="outdoor rugs,8x10 rugs,rug sizes,checkered rugs,5x7 rugs",
         title="Ruggable",
         company_url="https://ruggable.com",
-        serp_search_type=SerpSearchType.SEARCH,
-        serp_search_location=SerpSearchLocation.UNITED_STATES,
+        serp_search_type=SerpSearchType.search.name,
+        serp_search_location=SerpSearchLocations.UNITED_STATES.api_value,
         enable_html=False,
         selected_model=LargeLanguageModels.text_davinci_003.name,
         sampling_temperature=0.8,
@@ -85,9 +82,7 @@ class RequestModelBase(BasePage.RequestModel):
         task_instructions: str | None
         query_instructions: str | None
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
         max_search_urls: int | None
 
@@ -95,7 +90,7 @@ class RequestModelBase(BasePage.RequestModel):
         max_context_words: int | None
         scroll_jump: int | None
 
-        embedding_model: typing.Literal[tuple(e.name for e in EmbeddingModels)] | None
+        embedding_model: EmbeddingModels.api_enum | None
         dense_weight: float | None = DocSearchRequest.__fields__[
             "dense_weight"
         ].field_info
@@ -216,7 +211,7 @@ def run_v2(
         self,
         request: "GoogleGPTPage.RequestModel",
         response: "GoogleGPTPage.ResponseModel",
-    ):
+    ) -> typing.Iterator[str | None]:
         model = LargeLanguageModels[request.selected_model]
 
         query_instructions = (request.query_instructions or "").strip()
@@ -236,8 +231,8 @@ def run_v2(
             )
         response.serp_results, links = get_links_from_serp_api(
             response.final_search_query,
-            search_type=request.serp_search_type,
-            search_location=request.serp_search_location,
+            search_type=SerpSearchType.from_api(request.serp_search_type),
+            search_location=SerpSearchLocations.from_api(request.serp_search_location),
         )
         # extract links & their corresponding titles
         link_titles = {item.url: f"{item.title} | {item.snippet}" for item in links}
diff --git a/recipes/GoogleImageGen.py b/recipes/GoogleImageGen.py
index a94e812ce..61488b82d 100644
--- a/recipes/GoogleImageGen.py
+++ b/recipes/GoogleImageGen.py
@@ -17,14 +17,14 @@
 from daras_ai_v2.loom_video_widget import youtube_video
 from daras_ai_v2.serp_search import call_serp_api
 from daras_ai_v2.serp_search_locations import (
-    serp_search_location_selectbox,
     GoogleSearchLocationMixin,
+    SerpSearchLocations,
     SerpSearchType,
-    SerpSearchLocation,
+    serp_search_location_selectbox,
 )
 from daras_ai_v2.stable_diffusion import (
     img2img,
-    Img2ImgModels,
+    ImageToImageModels,
     SD_IMG_MAX_SIZE,
     instruct_pix2pix,
 )
@@ -37,6 +37,7 @@ class GoogleImageGenPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/eb23c078-88da-11ee-aa86-02420a000165/web%20search%20render.png.png"
     workflow = Workflow.GOOGLE_IMAGE_GEN
     slug_versions = ["GoogleImageGen", "render-images-with-ai"]
+    sdk_method_name = "imageFromWebSearch"
 
     sane_defaults = dict(
         num_outputs=1,
@@ -46,15 +47,15 @@ class GoogleImageGenPage(BasePage):
         sd_2_upscaling=False,
         seed=42,
         image_guidance_scale=1.2,
-        serp_search_type=SerpSearchType.SEARCH,
-        serp_search_location=SerpSearchLocation.UNITED_STATES,
+        serp_search_type=SerpSearchType.search.name,
+        serp_search_location=SerpSearchLocations.UNITED_STATES.api_value,
     )
 
     class RequestModel(GoogleSearchLocationMixin, BasePage.RequestModel):
         search_query: str
         text_prompt: str
 
-        selected_model: typing.Literal[tuple(e.name for e in Img2ImgModels)] | None
+        selected_model: ImageToImageModels.api_enum | None
 
         negative_prompt: str | None
 
@@ -112,8 +113,8 @@ def run(self, state: dict):
 
         serp_results = call_serp_api(
             request.search_query,
-            search_type=SerpSearchType.IMAGES,
-            search_location=request.serp_search_location,
+            search_type=SerpSearchType.images,
+            search_location=SerpSearchLocations.from_api(request.serp_search_location),
         )
         image_urls = [
             link
@@ -152,7 +153,7 @@ def run(self, state: dict):
 
         yield "Generating Images..."
 
-        if request.selected_model == Img2ImgModels.instruct_pix2pix.name:
+        if request.selected_model == ImageToImageModels.instruct_pix2pix.name:
             state["output_images"] = instruct_pix2pix(
                 prompt=request.text_prompt,
                 num_outputs=request.num_outputs,
@@ -185,7 +186,7 @@ def render_form_v2(self):
             """,
             key="search_query",
         )
-        model_selector(Img2ImgModels)
+        model_selector(ImageToImageModels)
         gui.text_area(
             """
             #### 👩‍💻 Prompt
@@ -199,7 +200,7 @@ def render_usage_guide(self):
         youtube_video("rnjvtaYYe8g")
 
     def render_settings(self):
-        img_model_settings(Img2ImgModels, render_model_selector=False)
+        img_model_settings(ImageToImageModels, render_model_selector=False)
         serp_search_location_selectbox()
 
     def render_output(self):
diff --git a/recipes/ImageSegmentation.py b/recipes/ImageSegmentation.py
index f886c2b4e..c9e16d988 100644
--- a/recipes/ImageSegmentation.py
+++ b/recipes/ImageSegmentation.py
@@ -16,7 +16,7 @@
 )
 from daras_ai_v2.base import BasePage
 from daras_ai_v2.enum_selector_widget import enum_selector
-from daras_ai_v2.image_segmentation import u2net, ImageSegmentationModels, dis
+from daras_ai_v2.image_segmentation import ImageSegmentationModels, dis, u2net
 from daras_ai_v2.img_io import opencv_to_pil, pil_to_bytes
 from daras_ai_v2.loom_video_widget import youtube_video
 from daras_ai_v2.polygon_fitter import (
@@ -37,6 +37,7 @@ class ImageSegmentationPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/06fc595e-88db-11ee-b428-02420a000168/AI%20Background%20Remover.png.png"
     workflow = Workflow.IMAGE_SEGMENTATION
     slug_versions = ["ImageSegmentation", "remove-image-background-with-ai"]
+    sdk_method_name = "removeBackground"
 
     sane_defaults = {
         "mask_threshold": 0.5,
@@ -50,9 +51,7 @@ class ImageSegmentationPage(BasePage):
     class RequestModel(BasePage.RequestModel):
         input_image: FieldHttpUrl
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in ImageSegmentationModels)] | None
-        )
+        selected_model: ImageSegmentationModels.api_enum | None
         mask_threshold: float | None
 
         rect_persepective_transform: bool | None
diff --git a/recipes/Img2Img.py b/recipes/Img2Img.py
index 97de89ab7..601a1d378 100644
--- a/recipes/Img2Img.py
+++ b/recipes/Img2Img.py
@@ -2,7 +2,7 @@
 
 from daras_ai_v2.pydantic_validation import FieldHttpUrl
 import requests
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
 import gooey_gui as gui
 from bots.models import Workflow
@@ -10,7 +10,7 @@
 from daras_ai_v2.img_model_settings_widgets import img_model_settings
 from daras_ai_v2.loom_video_widget import youtube_video
 from daras_ai_v2.stable_diffusion import (
-    Img2ImgModels,
+    ImageToImageModels,
     img2img,
     SD_IMG_MAX_SIZE,
     instruct_pix2pix,
@@ -27,6 +27,7 @@ class Img2ImgPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/bcc9351a-88d9-11ee-bf6c-02420a000166/Edit%20an%20image%20with%20AI%201.png.png"
     workflow = Workflow.IMG_2_IMG
     slug_versions = ["Img2Img", "ai-photo-editor"]
+    sdk_method_name = "remixImage"
 
     sane_defaults = {
         "num_outputs": 1,
@@ -45,12 +46,10 @@ class RequestModel(BasePage.RequestModel):
         input_image: FieldHttpUrl
         text_prompt: str | None
 
-        selected_model: typing.Literal[tuple(e.name for e in Img2ImgModels)] | None
+        selected_model: ImageToImageModels.api_enum | None
         selected_controlnet_model: (
-            list[typing.Literal[tuple(e.name for e in ControlNetModels)]]
-            | typing.Literal[tuple(e.name for e in ControlNetModels)]
-            | None
-        )
+            list[ControlNetModels.api_enum] | ControlNetModels.api_enum | None
+        ) = Field(**{"x-fern-type-name": "SelectedControlNetModels"})
         negative_prompt: str | None
 
         num_outputs: int | None
@@ -124,7 +123,7 @@ def render_description(self):
         )
 
     def render_settings(self):
-        img_model_settings(Img2ImgModels)
+        img_model_settings(ImageToImageModels)
 
     def render_usage_guide(self):
         youtube_video("narcZNyuNAg")
@@ -161,7 +160,7 @@ def run(self, state: dict) -> typing.Iterator[str | None]:
 
         yield "Generating Image..."
 
-        if request.selected_model == Img2ImgModels.instruct_pix2pix.name:
+        if request.selected_model == ImageToImageModels.instruct_pix2pix.name:
             state["output_images"] = instruct_pix2pix(
                 prompt=request.text_prompt,
                 num_outputs=request.num_outputs,
@@ -205,7 +204,7 @@ def preview_description(self, state: dict) -> str:
     def get_raw_price(self, state: dict) -> int:
         selected_model = state.get("selected_model")
         match selected_model:
-            case Img2ImgModels.dall_e.name:
+            case ImageToImageModels.dall_e.name:
                 unit_price = 20
             case _:
                 unit_price = 5
diff --git a/recipes/LetterWriter.py b/recipes/LetterWriter.py
index ff39fb3aa..53a0296af 100644
--- a/recipes/LetterWriter.py
+++ b/recipes/LetterWriter.py
@@ -18,6 +18,7 @@ class LetterWriterPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/aeb83ee8-889e-11ee-93dc-02420a000143/Youtube%20transcripts%20GPT%20extractions.png.png"
     workflow = Workflow.LETTER_WRITER
     slug_versions = ["LetterWriter"]
+    sdk_method_name = ""
 
     class RequestModel(BasePage.RequestModel):
         action_id: str
@@ -46,6 +47,10 @@ class ResponseModel(BaseModel):
         generated_input_prompt: str
         final_prompt: str
 
+    @classmethod
+    def get_openapi_extra(cls) -> dict[str, typing.Any]:
+        return {"x-fern-ignore": True}
+
     def render_description(self):
         gui.write(
             """
diff --git a/recipes/Lipsync.py b/recipes/Lipsync.py
index 6dae9320e..0a71e7ac8 100644
--- a/recipes/Lipsync.py
+++ b/recipes/Lipsync.py
@@ -8,7 +8,7 @@
 from daras_ai_v2.base import BasePage
 from daras_ai_v2.enum_selector_widget import enum_selector
 from daras_ai_v2.lipsync_api import run_wav2lip, run_sadtalker, LipsyncSettings
-from daras_ai_v2.lipsync_settings_widgets import lipsync_settings, LipsyncModel
+from daras_ai_v2.lipsync_settings_widgets import lipsync_settings, LipsyncModels
 from daras_ai_v2.loom_video_widget import youtube_video
 from daras_ai_v2.pydantic_validation import FieldHttpUrl
 
@@ -19,7 +19,7 @@
 
 
 def price_for_model(selected_model: str | None) -> float:
-    if selected_model == LipsyncModel.SadTalker.name:
+    if selected_model == LipsyncModels.SadTalker.name:
         multiplier = 2
     else:
         multiplier = 1
@@ -31,11 +31,10 @@ class LipsyncPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/f33e6332-88d8-11ee-89f9-02420a000169/Lipsync%20TTS.png.png"
     workflow = Workflow.LIPSYNC
     slug_versions = ["Lipsync"]
+    sdk_method_name = "lipsync"
 
     class RequestModel(LipsyncSettings, BasePage.RequestModel):
-        selected_model: typing.Literal[tuple(e.name for e in LipsyncModel)] = (
-            LipsyncModel.Wav2Lip.name
-        )
+        selected_model: LipsyncModels.api_enum = LipsyncModels.Wav2Lip.name
         input_audio: FieldHttpUrl = None
 
     class ResponseModel(BaseModel):
@@ -71,7 +70,7 @@ def render_form_v2(self):
             )
 
         enum_selector(
-            LipsyncModel,
+            LipsyncModels,
             label="###### Lipsync Model",
             key="selected_model",
             use_selectbox=True,
@@ -92,10 +91,10 @@ def run(self, state: dict) -> typing.Iterator[str | None]:
         else:
             max_frames = 250
 
-        model = LipsyncModel[request.selected_model]
+        model = LipsyncModels[request.selected_model]
         yield f"Running {model.value}..."
         match model:
-            case LipsyncModel.Wav2Lip:
+            case LipsyncModels.Wav2Lip:
                 state["output_video"], state["duration_sec"] = run_wav2lip(
                     face=request.input_face,
                     audio=request.input_audio,
@@ -107,7 +106,7 @@ def run(self, state: dict) -> typing.Iterator[str | None]:
                     ),
                     max_frames=max_frames,
                 )
-            case LipsyncModel.SadTalker:
+            case LipsyncModels.SadTalker:
                 state["output_video"], state["duration_sec"] = run_sadtalker(
                     request.sadtalker_settings,
                     face=request.input_face,
diff --git a/recipes/LipsyncTTS.py b/recipes/LipsyncTTS.py
index d557cd663..ef9f9b57d 100644
--- a/recipes/LipsyncTTS.py
+++ b/recipes/LipsyncTTS.py
@@ -6,7 +6,7 @@
 import gooey_gui as gui
 from bots.models import Workflow
 from daras_ai_v2.enum_selector_widget import enum_selector
-from daras_ai_v2.lipsync_api import LipsyncSettings, LipsyncModel
+from daras_ai_v2.lipsync_api import LipsyncSettings, LipsyncModels
 from daras_ai_v2.loom_video_widget import youtube_video
 from daras_ai_v2.safety_checker import safety_checker
 from daras_ai_v2.text_to_speech_settings_widgets import (
@@ -23,6 +23,7 @@ class LipsyncTTSPage(LipsyncPage, TextToSpeechPage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/1acfa370-88d9-11ee-bf6c-02420a000166/Lipsync%20with%20audio%201.png.png"
     workflow = Workflow.LIPSYNC_TTS
     slug_versions = ["LipsyncTTS", "lipsync-maker"]
+    sdk_method_name = "lipsyncTTS"
 
     sane_defaults = {
         "elevenlabs_model": "eleven_multilingual_v2",
@@ -31,9 +32,7 @@ class LipsyncTTSPage(LipsyncPage, TextToSpeechPage):
     }
 
     class RequestModel(LipsyncSettings, TextToSpeechPage.RequestModel):
-        selected_model: typing.Literal[tuple(e.name for e in LipsyncModel)] = (
-            LipsyncModel.Wav2Lip.name
-        )
+        selected_model: LipsyncModels.api_enum = LipsyncModels.Wav2Lip.name
 
     class ResponseModel(BaseModel):
         audio_url: str | None
@@ -76,7 +75,7 @@ def render_form_v2(self):
             )
 
         enum_selector(
-            LipsyncModel,
+            LipsyncModels,
             label="###### Lipsync Model",
             key="selected_model",
             use_selectbox=True,
diff --git a/recipes/ObjectInpainting.py b/recipes/ObjectInpainting.py
index 3ec1f6f89..65f902ca4 100644
--- a/recipes/ObjectInpainting.py
+++ b/recipes/ObjectInpainting.py
@@ -33,6 +33,7 @@ class ObjectInpaintingPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/f07b731e-88d9-11ee-a658-02420a000163/W.I.3.png.png"
     workflow = Workflow.OBJECT_INPAINTING
     slug_versions = ["ObjectInpainting", "product-photo-background-generator"]
+    sdk_method_name = "productImage"
 
     sane_defaults = {
         "mask_threshold": 0.7,
@@ -55,7 +56,7 @@ class RequestModel(BasePage.RequestModel):
 
         mask_threshold: float | None
 
-        selected_model: typing.Literal[tuple(e.name for e in InpaintingModels)] | None
+        selected_model: InpaintingModels.api_enum | None
 
         negative_prompt: str | None
 
diff --git a/recipes/QRCodeGenerator.py b/recipes/QRCodeGenerator.py
index cfa9ce013..7f61f8d76 100644
--- a/recipes/QRCodeGenerator.py
+++ b/recipes/QRCodeGenerator.py
@@ -8,7 +8,7 @@
 from django.core.exceptions import ValidationError
 from django.core.validators import URLValidator
 from furl import furl
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 from pyzbar import pyzbar
 
 import gooey_gui as gui
@@ -30,14 +30,13 @@
 from daras_ai_v2.repositioning import reposition_object, repositioning_preview_widget
 from daras_ai_v2.safety_checker import safety_checker
 from daras_ai_v2.stable_diffusion import (
-    Text2ImgModels,
+    TextToImageModels,
     controlnet,
     ControlNetModels,
-    Img2ImgModels,
+    ImageToImageModels,
     Schedulers,
 )
 from daras_ai_v2.vcard import VCARD
-from recipes.EmailFaceInpainting import get_photo_for_email
 from recipes.SocialLookupEmail import get_profile_for_email
 from url_shortener.models import ShortenedURL
 from daras_ai_v2.enum_selector_widget import enum_multiselect
@@ -58,6 +57,7 @@ class QRCodeGeneratorPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/03d6538e-88d5-11ee-ad97-02420a00016c/W.I.2.png.png"
     workflow = Workflow.QR_CODE
     slug_versions = ["art-qr-code", "qr", "qr-code"]
+    sdk_method_name = "qrCode"
 
     sane_defaults = dict(
         num_outputs=2,
@@ -82,7 +82,9 @@ def __init__(self, *args, **kwargs):
     class RequestModel(BasePage.RequestModel):
         qr_code_data: str | None
         qr_code_input_image: FieldHttpUrl | None
-        qr_code_vcard: VCARD | None
+        qr_code_vcard: VCARD | None = Field(
+            title="VCard", **{"x-fern-type-name": "VCard"}
+        )
         qr_code_file: FieldHttpUrl | None
 
         use_url_shortener: bool | None
@@ -90,18 +92,14 @@ class RequestModel(BasePage.RequestModel):
         text_prompt: str
         negative_prompt: str | None
         image_prompt: str | None
-        image_prompt_controlnet_models: (
-            list[typing.Literal[tuple(e.name for e in ControlNetModels)], ...] | None
-        )
+        image_prompt_controlnet_models: list[ControlNetModels.api_enum] | None
         image_prompt_strength: float | None
         image_prompt_scale: float | None
         image_prompt_pos_x: float | None
         image_prompt_pos_y: float | None
 
-        selected_model: typing.Literal[tuple(e.name for e in Text2ImgModels)] | None
-        selected_controlnet_model: (
-            list[typing.Literal[tuple(e.name for e in ControlNetModels)], ...] | None
-        )
+        selected_model: TextToImageModels.api_enum | None
+        selected_controlnet_model: list[ControlNetModels.api_enum] | None
 
         output_width: int | None
         output_height: int | None
@@ -111,7 +109,7 @@ class RequestModel(BasePage.RequestModel):
 
         num_outputs: int | None
         quality: int | None
-        scheduler: typing.Literal[tuple(e.name for e in Schedulers)] | None
+        scheduler: Schedulers.api_enum | None
 
         seed: int | None
 
@@ -297,7 +295,7 @@ def render_settings(self):
         )
 
         img_model_settings(
-            Img2ImgModels,
+            ImageToImageModels,
             show_scheduler=True,
             require_controlnet=True,
             extra_explanations={
@@ -486,7 +484,7 @@ def run(self, state: dict) -> typing.Iterator[str | None]:
 
         state["raw_images"] = raw_images = []
 
-        yield f"Running {Text2ImgModels[request.selected_model].value}..."
+        yield f"Running {TextToImageModels[request.selected_model].label}..."
         if isinstance(request.selected_controlnet_model, str):
             request.selected_controlnet_model = [request.selected_controlnet_model]
         init_images = [image] * len(request.selected_controlnet_model)
@@ -547,12 +545,14 @@ def preview_description(self, state: dict) -> str:
         """
 
     def get_raw_price(self, state: dict) -> int:
-        selected_model = state.get("selected_model", Text2ImgModels.dream_shaper.name)
+        selected_model = state.get(
+            "selected_model", TextToImageModels.dream_shaper.name
+        )
         total = 5
         match selected_model:
-            case Text2ImgModels.deepfloyd_if.name:
+            case TextToImageModels.deepfloyd_if.name:
                 total += 3
-            case Text2ImgModels.dall_e.name:
+            case TextToImageModels.dall_e.name:
                 total += 10
         return total * state.get("num_outputs", 1)
 
diff --git a/recipes/RelatedQnA.py b/recipes/RelatedQnA.py
index 6372ce65e..f43c84097 100644
--- a/recipes/RelatedQnA.py
+++ b/recipes/RelatedQnA.py
@@ -8,10 +8,7 @@
     LargeLanguageModels,
 )
 from daras_ai_v2.serp_search import get_related_questions_from_serp_api
-from daras_ai_v2.serp_search_locations import (
-    SerpSearchLocation,
-    SerpSearchType,
-)
+from daras_ai_v2.serp_search_locations import SerpSearchLocations, SerpSearchType
 from recipes.DocSearch import render_doc_search_step, EmptySearchResults
 from recipes.GoogleGPT import GoogleGPTPage
 from recipes.RelatedQnADoc import render_qna_outputs
@@ -28,6 +25,7 @@ class RelatedQnAPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/37b0ba22-88d6-11ee-b549-02420a000167/People%20also%20ask.png.png"
     workflow = Workflow.RELATED_QNA_MAKER
     slug_versions = ["related-qna-maker"]
+    sdk_method_name = "seoPeopleAlsoAsk"
 
     price = 75
 
@@ -36,8 +34,8 @@ class RelatedQnAPage(BasePage):
         max_context_words=200,
         scroll_jump=5,
         dense_weight=1.0,
-        serp_search_type=SerpSearchType.SEARCH,
-        serp_search_location=SerpSearchLocation.UNITED_STATES,
+        serp_search_type=SerpSearchType.search.name,
+        serp_search_location=SerpSearchLocations.UNITED_STATES.api_value,
     )
 
     class RequestModel(GoogleGPTPage.RequestModel):
@@ -117,7 +115,7 @@ def run_v2(
             related_questions,
         ) = get_related_questions_from_serp_api(
             request.search_query,
-            search_location=request.serp_search_location,
+            search_location=SerpSearchLocations.from_api(request.serp_search_location),
         )
 
         all_questions = [request.search_query] + related_questions[:9]
diff --git a/recipes/RelatedQnADoc.py b/recipes/RelatedQnADoc.py
index 3f8c2d2d8..8362a7046 100644
--- a/recipes/RelatedQnADoc.py
+++ b/recipes/RelatedQnADoc.py
@@ -7,10 +7,7 @@
 from daras_ai_v2.language_model import LargeLanguageModels
 from daras_ai_v2.search_ref import CitationStyles
 from daras_ai_v2.serp_search import get_related_questions_from_serp_api
-from daras_ai_v2.serp_search_locations import (
-    SerpSearchLocation,
-    SerpSearchType,
-)
+from daras_ai_v2.serp_search_locations import SerpSearchLocations, SerpSearchType
 from daras_ai_v2.vector_search import render_sources_widget
 from recipes.DocSearch import DocSearchPage, render_doc_search_step, EmptySearchResults
 from recipes.GoogleGPT import render_output_with_refs, GoogleSearchMixin
@@ -27,14 +24,15 @@ class RelatedQnADocPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/aeb83ee8-889e-11ee-93dc-02420a000143/Youtube%20transcripts%20GPT%20extractions.png.png"
     workflow = Workflow.RELATED_QNA_MAKER_DOC
     slug_versions = ["related-qna-maker-doc"]
+    sdk_method_name = "seoPeopleAlsoAskDoc"
 
     price = 100
 
     sane_defaults = dict(
         citation_style=CitationStyles.number.name,
         dense_weight=1.0,
-        serp_search_type=SerpSearchType.SEARCH,
-        serp_search_location=SerpSearchLocation.UNITED_STATES,
+        serp_search_type=SerpSearchType.search.name,
+        serp_search_location=SerpSearchLocations.UNITED_STATES.api_value,
     )
 
     class RequestModel(GoogleSearchMixin, DocSearchPage.RequestModel):
@@ -111,7 +109,7 @@ def run_v2(
             related_questions,
         ) = get_related_questions_from_serp_api(
             request.search_query,
-            search_location=request.serp_search_location,
+            search_location=SerpSearchLocations.from_api(request.serp_search_location),
         )
 
         all_questions = [request.search_query] + related_questions[:9]
diff --git a/recipes/SEOSummary.py b/recipes/SEOSummary.py
index 1b4e65365..2274b4662 100644
--- a/recipes/SEOSummary.py
+++ b/recipes/SEOSummary.py
@@ -28,9 +28,9 @@
 from daras_ai_v2.scrollable_html_widget import scrollable_html
 from daras_ai_v2.serp_search import get_links_from_serp_api
 from daras_ai_v2.serp_search_locations import (
-    serp_search_settings,
-    SerpSearchLocation,
+    SerpSearchLocations,
     SerpSearchType,
+    serp_search_settings,
 )
 from daras_ai_v2.settings import EXTERNAL_REQUEST_TIMEOUT_SEC
 from recipes.GoogleGPT import GoogleSearchMixin
@@ -61,6 +61,7 @@ class SEOSummaryPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/85f38b42-88d6-11ee-ad97-02420a00016c/Create%20SEO%20optimized%20content%20option%202.png.png"
     workflow = Workflow.SEO_SUMMARY
     slug_versions = ["SEOSummary", "seo-paragraph-generator"]
+    sdk_method_name = "seoContent"
 
     def preview_image(self, state: dict) -> str | None:
         return SEO_SUMMARY_DEFAULT_META_IMG
@@ -73,8 +74,8 @@ def preview_description(self, state: dict) -> str:
         keywords="outdoor rugs,8x10 rugs,rug sizes,checkered rugs,5x7 rugs",
         title="Ruggable",
         company_url="https://ruggable.com",
-        serp_search_type=SerpSearchType.SEARCH,
-        serp_search_location=SerpSearchLocation.UNITED_STATES,
+        serp_search_type=SerpSearchType.search.name,
+        serp_search_location=SerpSearchLocations.UNITED_STATES.api_value,
         enable_html=False,
         selected_model=LargeLanguageModels.text_davinci_003.name,
         sampling_temperature=0.8,
@@ -90,6 +91,9 @@ def preview_description(self, state: dict) -> str:
     )
 
     class RequestModelBase(BaseModel):
+        class Config:
+            use_enum_values = True
+
         search_query: str
         keywords: str
         title: str
@@ -99,9 +103,7 @@ class RequestModelBase(BaseModel):
 
         enable_html: bool | None
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
         max_search_urls: int | None
 
@@ -274,8 +276,8 @@ def run(self, state: dict) -> typing.Iterator[str | None]:
 
         serp_results, links = get_links_from_serp_api(
             request.search_query,
-            search_type=request.serp_search_type,
-            search_location=request.serp_search_location,
+            search_type=SerpSearchType.from_api(request.serp_search_type),
+            search_location=SerpSearchLocations.from_api(request.serp_search_location),
         )
         state["serp_results"] = serp_results
         state["search_urls"] = [it.url for it in links]
@@ -314,8 +316,8 @@ def _crosslink_keywords(output_content, request):
     all_results = map_parallel(
         lambda keyword: get_links_from_serp_api(
             f"site:{host} {keyword}",
-            search_type=request.serp_search_type,
-            search_location=request.serp_search_location,
+            search_type=SerpSearchType.from_api(request.serp_search_type),
+            search_location=SerpSearchLocations.from_api(request.serp_search_location),
         )[1],
         relevant_keywords,
     )
diff --git a/recipes/SmartGPT.py b/recipes/SmartGPT.py
index fa4066fc7..7141132c8 100644
--- a/recipes/SmartGPT.py
+++ b/recipes/SmartGPT.py
@@ -1,7 +1,7 @@
 import typing
 
 import jinja2.sandbox
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
 import gooey_gui as gui
 from bots.models import Workflow
@@ -29,6 +29,7 @@ class SmartGPTPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/ffd24ad8-88d7-11ee-a658-02420a000163/SmartGPT.png.png"
     workflow = Workflow.SMART_GPT
     slug_versions = ["SmartGPT"]
+    sdk_method_name = "smartGPT"
     price = 20
 
     class RequestModelBase(BasePage.RequestModel):
@@ -38,9 +39,7 @@ class RequestModelBase(BasePage.RequestModel):
         reflexion_prompt: str | None
         dera_prompt: str | None
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
     class RequestModel(LanguageModelSettings, RequestModelBase):
         pass
diff --git a/recipes/SocialLookupEmail.py b/recipes/SocialLookupEmail.py
index bc3a0dea1..f1e3a3bd5 100644
--- a/recipes/SocialLookupEmail.py
+++ b/recipes/SocialLookupEmail.py
@@ -2,7 +2,7 @@
 import typing
 
 import requests
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
 import gooey_gui as gui
 from bots.models import Workflow
@@ -32,6 +32,7 @@ class SocialLookupEmailPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/5fbd475a-88d7-11ee-aac9-02420a00016b/personalized%20email.png.png"
     workflow = Workflow.SOCIAL_LOOKUP_EMAIL
     slug_versions = ["SocialLookupEmail", "email-writer-with-profile-lookup"]
+    sdk_method_name = "personalizeEmail"
 
     sane_defaults = {
         "selected_model": LargeLanguageModels.gpt_4.name,
@@ -54,9 +55,7 @@ class RequestModelBase(BasePage.RequestModel):
         # domain: str | None
         # key_words: str | None
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
 
     class RequestModel(LanguageModelSettings, RequestModelBase):
         pass
diff --git a/recipes/Text2Audio.py b/recipes/Text2Audio.py
index f3199a95b..302bf236f 100644
--- a/recipes/Text2Audio.py
+++ b/recipes/Text2Audio.py
@@ -1,6 +1,6 @@
 import typing
-from enum import Enum
 
+from daras_ai_v2.custom_enum import GooeyEnum
 from daras_ai_v2.pydantic_validation import FieldHttpUrl
 from pydantic import BaseModel
 
@@ -18,7 +18,7 @@
 DEFAULT_TEXT2AUDIO_META_IMG = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/85cf8ea4-9457-11ee-bd77-02420a0001ce/Text%20guided%20audio.jpg.png"
 
 
-class Text2AudioModels(Enum):
+class Text2AudioModels(GooeyEnum):
     audio_ldm = "AudioLDM (CVSSP)"
 
 
@@ -32,6 +32,7 @@ class Text2AudioPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/a4481d58-88d9-11ee-aa86-02420a000165/Text%20guided%20audio%20generator.png.png"
     workflow = Workflow.TEXT_2_AUDIO
     slug_versions = ["text2audio"]
+    sdk_method_name = "textToMusic"
 
     sane_defaults = dict(
         seed=42,
@@ -50,13 +51,11 @@ class RequestModel(BasePage.RequestModel):
         seed: int | None
         sd_2_upscaling: bool | None
 
-        selected_models: (
-            list[typing.Literal[tuple(e.name for e in Text2AudioModels)]] | None
-        )
+        selected_models: list[Text2AudioModels.api_enum] | None
 
     class ResponseModel(BaseModel):
         output_audios: dict[
-            typing.Literal[tuple(e.name for e in Text2AudioModels)],
+            Text2AudioModels.api_enum,
             list[FieldHttpUrl],
         ]
 
diff --git a/recipes/TextToSpeech.py b/recipes/TextToSpeech.py
index dbc877a17..0488a6fba 100644
--- a/recipes/TextToSpeech.py
+++ b/recipes/TextToSpeech.py
@@ -31,7 +31,7 @@
 
 
 class TextToSpeechSettings(BaseModel):
-    tts_provider: typing.Literal[tuple(e.name for e in TextToSpeechProviders)] | None
+    tts_provider: TextToSpeechProviders.api_enum | None
 
     uberduck_voice_name: str | None
     uberduck_speaking_rate: float | None
@@ -55,8 +55,8 @@ class TextToSpeechSettings(BaseModel):
 
     azure_voice_name: str | None
 
-    openai_voice_name: OpenAI_TTS_Voices.api_choices | None
-    openai_tts_model: OpenAI_TTS_Models.api_choices | None
+    openai_voice_name: OpenAI_TTS_Voices.api_enum | None
+    openai_tts_model: OpenAI_TTS_Models.api_enum | None
 
 
 class TextToSpeechPage(BasePage):
@@ -69,6 +69,7 @@ class TextToSpeechPage(BasePage):
         "text2speech",
         "compare-text-to-speech-engines",
     ]
+    sdk_method_name = "textToSpeech"
 
     sane_defaults = {
         "tts_provider": TextToSpeechProviders.GOOGLE_TTS.value,
diff --git a/recipes/Translation.py b/recipes/Translation.py
index b8061beb2..19ce68795 100644
--- a/recipes/Translation.py
+++ b/recipes/Translation.py
@@ -38,13 +38,14 @@ class TranslationPage(BasePage):
     title = "Compare AI Translations"
     workflow = Workflow.TRANSLATION
     slug_versions = ["translate", "translation", "compare-ai-translation"]
+    sdk_method_name = "translate"
 
     class RequestModelBase(BasePage.RequestModel):
         texts: list[str] = Field([])
 
-        selected_model: (
-            typing.Literal[tuple(e.name for e in TranslationModels)]
-        ) | None = Field(TranslationModels.google.name)
+        selected_model: TranslationModels.api_enum | None = Field(
+            TranslationModels.google.name
+        )
 
     class RequestModel(TranslationOptions, RequestModelBase):
         pass
diff --git a/recipes/VideoBots.py b/recipes/VideoBots.py
index 77b266c26..e6637c0d3 100644
--- a/recipes/VideoBots.py
+++ b/recipes/VideoBots.py
@@ -79,7 +79,7 @@
     language_model_selector,
     LanguageModelSettings,
 )
-from daras_ai_v2.lipsync_api import LipsyncSettings, LipsyncModel
+from daras_ai_v2.lipsync_api import LipsyncSettings, LipsyncModels
 from daras_ai_v2.lipsync_settings_widgets import lipsync_settings
 from daras_ai_v2.loom_video_widget import youtube_video
 from daras_ai_v2.prompt_vars import render_prompt_vars
@@ -136,6 +136,9 @@ class VideoBotsPage(BasePage):
     workflow = Workflow.VIDEO_BOTS
     slug_versions = ["video-bots", "bots", "copilot"]
 
+    sdk_group_name = "copilot"
+    sdk_method_name = "completion"
+
     functions_in_settings = False
 
     sane_defaults = {
@@ -193,9 +196,7 @@ class RequestModelBase(BasePage.RequestModel):
         bot_script: str | None
 
         # llm model
-        selected_model: (
-            typing.Literal[tuple(e.name for e in LargeLanguageModels)] | None
-        )
+        selected_model: LargeLanguageModels.api_enum | None
         document_model: str | None = Field(
             title="🩻 Photo / Document Intelligence",
             description="When your copilot users upload a photo or pdf, what kind of document are they mostly likely to upload? "
@@ -211,15 +212,15 @@ class RequestModelBase(BasePage.RequestModel):
         max_context_words: int | None
         scroll_jump: int | None
 
-        embedding_model: typing.Literal[tuple(e.name for e in EmbeddingModels)] | None
+        embedding_model: EmbeddingModels.api_enum | None
         dense_weight: float | None = DocSearchRequest.__fields__[
             "dense_weight"
         ].field_info
 
-        citation_style: typing.Literal[tuple(e.name for e in CitationStyles)] | None
+        citation_style: CitationStyles.api_enum | None
         use_url_shortener: bool | None
 
-        asr_model: typing.Literal[tuple(e.name for e in AsrModels)] | None = Field(
+        asr_model: AsrModels.api_enum | None = Field(
             title="Speech-to-Text Provider",
             description="Choose a model to transcribe incoming audio messages to text.",
         )
@@ -228,9 +229,7 @@ class RequestModelBase(BasePage.RequestModel):
             description="Choose a language to transcribe incoming audio messages to text.",
         )
 
-        translation_model: (
-            typing.Literal[tuple(e.name for e in TranslationModels)] | None
-        )
+        translation_model: TranslationModels.api_enum | None
         user_language: str | None = Field(
             title="User Language",
             description="Choose a language to translate incoming text & audio messages to English and responses back to your selected language. Useful for low-resource languages.",
@@ -249,9 +248,7 @@ class RequestModelBase(BasePage.RequestModel):
             """,
         )
 
-        lipsync_model: typing.Literal[tuple(e.name for e in LipsyncModel)] = (
-            LipsyncModel.Wav2Lip.name
-        )
+        lipsync_model: LipsyncModels.api_enum = LipsyncModels.Wav2Lip.name
 
         tools: list[LLMTools] | None = Field(
             title="🛠️ Tools",
@@ -286,6 +283,13 @@ class ResponseModel(BaseModel):
 
         finish_reason: list[str] | None
 
+    @classmethod
+    def get_openapi_extra(cls) -> dict[str, typing.Any]:
+        return {
+            "x-fern-sdk-group-name": cls.sdk_group_name,
+            "x-fern-sdk-method-name": cls.sdk_method_name,
+        }
+
     def preview_image(self, state: dict) -> str | None:
         return DEFAULT_COPILOT_META_IMG
 
@@ -372,7 +376,7 @@ def render_form_v2(self):
                 key="input_face",
             )
             enum_selector(
-                LipsyncModel,
+                LipsyncModels,
                 label="###### Lipsync Model",
                 key="lipsync_model",
                 use_selectbox=True,
diff --git a/recipes/asr_page.py b/recipes/asr_page.py
index f68be1a08..509540b5a 100644
--- a/recipes/asr_page.py
+++ b/recipes/asr_page.py
@@ -38,19 +38,18 @@ class AsrPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/5fb7e5f6-88d9-11ee-aa86-02420a000165/Speech.png.png"
     workflow = Workflow.ASR
     slug_versions = ["asr", "speech"]
+    sdk_method_name = "speechRecognition"
 
     sane_defaults = dict(output_format=AsrOutputFormat.text.name)
 
     class RequestModelBase(BasePage.RequestModel):
         documents: list[FieldHttpUrl]
-        selected_model: typing.Literal[tuple(e.name for e in AsrModels)] | None
+        selected_model: AsrModels.api_enum | None
         language: str | None
 
-        translation_model: (
-            typing.Literal[tuple(e.name for e in TranslationModels)] | None
-        )
+        translation_model: TranslationModels.api_enum | None
 
-        output_format: typing.Literal[tuple(e.name for e in AsrOutputFormat)] | None
+        output_format: AsrOutputFormat.api_enum | None
 
         google_translate_target: str | None = Field(
             deprecated=True,
diff --git a/recipes/embeddings_page.py b/recipes/embeddings_page.py
index e65580681..d3cff0244 100644
--- a/recipes/embeddings_page.py
+++ b/recipes/embeddings_page.py
@@ -17,11 +17,12 @@ class EmbeddingsPage(BasePage):
     explore_image = "https://storage.googleapis.com/dara-c1b52.appspot.com/daras_ai/media/aeb83ee8-889e-11ee-93dc-02420a000143/Youtube%20transcripts%20GPT%20extractions.png.png"
     workflow = Workflow.EMBEDDINGS
     slug_versions = ["embeddings", "embed", "text-embedings"]
+    sdk_method_name = "embed"
     price = 1
 
     class RequestModel(BasePage.RequestModel):
         texts: list[str]
-        selected_model: typing.Literal[tuple(e.name for e in EmbeddingModels)] | None
+        selected_model: EmbeddingModels.api_enum | None
 
     class ResponseModel(BaseModel):
         embeddings: list[list[float]]
diff --git a/routers/api.py b/routers/api.py
index 73c5ec7b9..02b8e357a 100644
--- a/routers/api.py
+++ b/routers/api.py
@@ -5,6 +5,7 @@
 import typing
 
 import gooey_gui as gui
+from fastapi import Query
 from fastapi import Depends
 from fastapi import Form
 from fastapi import HTTPException
@@ -152,10 +153,13 @@ def script_to_api(page_cls: typing.Type[BasePage]):
         operation_id=page_cls.slug_versions[0],
         tags=[page_cls.title],
         name=page_cls.title + " (v2 sync)",
+        openapi_extra={"x-fern-ignore": True},
+        include_in_schema=False,
     )
     def run_api_json(
         request: Request,
         page_request: request_model,
+        example_id: str | None = None,
         user: AppUser = Depends(api_auth_header),
     ):
         result, sr = submit_api_call(
@@ -200,11 +204,13 @@ def run_api_form(
         name=page_cls.title + " (v3 async)",
         tags=[page_cls.title],
         status_code=202,
+        openapi_extra=page_cls.get_openapi_extra(),
     )
     def run_api_json_async(
         request: Request,
         response: Response,
         page_request: request_model,
+        example_id: str | None = Query(default=None),
         user: AppUser = Depends(api_auth_header),
     ):
         result, sr = submit_api_call(
@@ -255,6 +261,7 @@ def run_api_form_async(
         operation_id="status__" + page_cls.slug_versions[0],
         tags=[page_cls.title],
         name=page_cls.title + " (v3 status)",
+        openapi_extra={"x-fern-ignore": True},
     )
     def get_run_status(
         run_id: str,
@@ -426,11 +433,16 @@ class BalanceResponse(BaseModel):
     balance: int = Field(description="Current balance in credits")
 
 
-@app.get("/v1/balance/", response_model=BalanceResponse, tags=["Misc"])
+@app.get(
+    "/v1/balance/",
+    response_model=BalanceResponse,
+    tags=["Misc"],
+    openapi_extra={"x-fern-sdk-method-name": "getBalance"},
+)
 def get_balance(user: AppUser = Depends(api_auth_header)):
     return BalanceResponse(balance=user.balance)
 
 
-@app.get("/status")
+@app.get("/status", openapi_extra={"x-fern-ignore": True})
 async def health():
     return "OK"
diff --git a/routers/bots_api.py b/routers/bots_api.py
index 021caae4b..c99dadec0 100644
--- a/routers/bots_api.py
+++ b/routers/bots_api.py
@@ -85,6 +85,7 @@ class CreateStreamResponse(BaseModel):
     operation_id=VideoBotsPage.slug_versions[0] + "__stream_create",
     tags=["Copilot Integrations"],
     name="Copilot Integrations Create Stream",
+    openapi_extra={"x-fern-ignore": True},
 )
 def stream_create(request: CreateStreamRequest, response: Response):
     request_id = str(uuid.uuid4())
@@ -173,6 +174,7 @@ class StreamError(BaseModel):
     operation_id=VideoBotsPage.slug_versions[0] + "__stream",
     tags=["Copilot Integrations"],
     name="Copilot integrations Stream Response",
+    openapi_extra={"x-fern-ignore": True},
 )
 def stream_response(request_id: str):
     r = get_redis_cache().getdel(f"gooey/stream-init/v1/{request_id}")
diff --git a/routers/broadcast_api.py b/routers/broadcast_api.py
index cbf1ca07c..82cf8ea59 100644
--- a/routers/broadcast_api.py
+++ b/routers/broadcast_api.py
@@ -51,6 +51,7 @@ class BotBroadcastRequestModel(BaseModel):
     operation_id=VideoBotsPage.slug_versions[0] + "__broadcast",
     tags=["Misc"],
     name=f"Send Broadcast Message",
+    openapi_extra={"x-fern-ignore": True},
 )
 @app.post(
     f"/v2/{VideoBotsPage.slug_versions[0]}/broadcast/send",
diff --git a/scripts/init_self_hosted_pricing.py b/scripts/init_self_hosted_pricing.py
index 68bd63512..4cb5ba224 100644
--- a/scripts/init_self_hosted_pricing.py
+++ b/scripts/init_self_hosted_pricing.py
@@ -2,12 +2,9 @@
 
 from daras_ai_v2.gpu_server import build_queue_name
 from daras_ai_v2.stable_diffusion import (
-    Text2ImgModels,
-    Img2ImgModels,
+    TextToImageModels,
+    ImageToImageModels,
     InpaintingModels,
-    text2img_model_ids,
-    img2img_model_ids,
-    inpaint_model_ids,
 )
 from recipes.DeforumSD import AnimationModels
 from usage_costs.models import ModelPricing
@@ -17,21 +14,15 @@
 
 
 def run():
-    for model in AnimationModels:
-        add_model(model.value, model.name)
-    for model_enum, model_ids in [
-        (Text2ImgModels, text2img_model_ids),
-        (Img2ImgModels, img2img_model_ids),
-        (InpaintingModels, inpaint_model_ids),
+    for model_enum in [
+        AnimationModels,
+        TextToImageModels,
+        ImageToImageModels,
+        InpaintingModels,
     ]:
         for m in model_enum:
-            if "dall_e" in m.name:
-                continue
-            try:
-                add_model(model_ids[m], m.name)
-            except KeyError:
-                pass
-
+            if "dall_e" not in m.name and m.model_id:
+                add_model(m.model_id, m.name)
     add_model("wav2lip_gan.pth", "wav2lip")
     add_model("SadTalker_V0.0.2_512.safetensors", "sadtalker")
 
diff --git a/scripts/run_all_diffusion.py b/scripts/run_all_diffusion.py
index 183260c5d..c6132b953 100644
--- a/scripts/run_all_diffusion.py
+++ b/scripts/run_all_diffusion.py
@@ -20,9 +20,9 @@
 from daras_ai_v2.stable_diffusion import (
     controlnet,
     ControlNetModels,
-    Img2ImgModels,
+    ImageToImageModels,
     text2img,
-    Text2ImgModels,
+    TextToImageModels,
     img2img,
     instruct_pix2pix,
     sd_upscale,
@@ -34,7 +34,7 @@
 
 # def fn():
 #     text2img(
-#         selected_model=Img2ImgModels.sd_1_5.name,
+#         selected_model=ImageToImageModels.sd_1_5.name,
 #         prompt=get_random_string(100, string.ascii_letters),
 #         num_outputs=1,
 #         num_inference_steps=1,
@@ -45,7 +45,7 @@
 #     # r = requests.get(GpuEndpoints.sd_multi / "magic")
 #     # raise_for_status(r)
 #     # img2img(
-#     #     selected_model=Img2ImgModels.sd_1_5.name,
+#     #     selected_model=ImageToImageModels.sd_1_5.name,
 #     #     prompt=get_random_string(100, string.ascii_letters),
 #     #     num_outputs=1,
 #     #     init_image=random_img,
@@ -55,7 +55,7 @@
 #     # )
 #     # controlnet(
 #     #     selected_controlnet_model=ControlNetModels.sd_controlnet_depth.name,
-#     #     selected_model=Img2ImgModels.sd_1_5.name,
+#     #     selected_model=ImageToImageModels.sd_1_5.name,
 #     #     prompt=get_random_string(100, string.ascii_letters),
 #     #     num_outputs=1,
 #     #     init_image=random_img,
@@ -72,11 +72,11 @@
 # exit()
 tasks = []
 
-for model in Img2ImgModels:
+for model in ImageToImageModels:
     if model in [
-        Img2ImgModels.instruct_pix2pix,
-        Img2ImgModels.dall_e,
-        Img2ImgModels.jack_qiao,
+        ImageToImageModels.instruct_pix2pix,
+        ImageToImageModels.dall_e,
+        ImageToImageModels.jack_qiao,
     ]:
         continue
     print(model)
@@ -96,7 +96,7 @@
     )
     for controlnet_model in ControlNetModels:
         if model in [
-            Img2ImgModels.sd_2,
+            ImageToImageModels.sd_2,
         ]:
             continue
         print(controlnet_model)
@@ -115,10 +115,10 @@
             )
         )
 
-for model in Text2ImgModels:
+for model in TextToImageModels:
     if model in [
-        Text2ImgModels.dall_e,
-        Text2ImgModels.jack_qiao,
+        TextToImageModels.dall_e,
+        TextToImageModels.jack_qiao,
     ]:
         continue
     print(model)
diff --git a/usage_costs/models.py b/usage_costs/models.py
index 8941d1cd7..2d113e2f9 100644
--- a/usage_costs/models.py
+++ b/usage_costs/models.py
@@ -58,14 +58,14 @@ class ModelProvider(models.IntegerChoices):
 def get_model_choices():
     from daras_ai_v2.language_model import LargeLanguageModels
     from recipes.DeforumSD import AnimationModels
-    from daras_ai_v2.stable_diffusion import Text2ImgModels, Img2ImgModels
+    from daras_ai_v2.stable_diffusion import TextToImageModels, ImageToImageModels
 
     return (
         [(api.name, api.value) for api in LargeLanguageModels]
         + [(model.name, model.label) for model in AnimationModels]
-        + [(model.name, model.value) for model in Text2ImgModels]
-        + [(model.name, model.value) for model in Img2ImgModels]
-        + [(model.name, model.value) for model in InpaintingModels]
+        + [(model.name, model.label) for model in TextToImageModels]
+        + [(model.name, model.label) for model in ImageToImageModels]
+        + [(model.name, model.label) for model in InpaintingModels]
         + [("wav2lip", "LipSync (wav2lip)")]
         + [("sadtalker", "LipSync (sadtalker)")]
     )