[pull] master from Comfy-Org:master

comfy/image_encoders/dino3.py

@@ -3,6 +3,7 @@
 import torch.nn as nn
 import torch.nn.functional as F
 
+import comfy.ops
 from comfy.ldm.modules.attention import optimized_attention_for_device
 from comfy.image_encoders.dino2 import LayerScale as DINOv3ViTLayerScale
 
@@ -171,11 +172,11 @@ def forward(self, pixel_values, bool_masked_pos=None):
         patch_embeddings = patch_embeddings.flatten(2).transpose(1, 2)
 
         if bool_masked_pos is not None:
-            mask_token = self.mask_token.to(patch_embeddings.dtype)
+            mask_token = comfy.ops.cast_to_input(self.mask_token, patch_embeddings)
             patch_embeddings = torch.where(bool_masked_pos.unsqueeze(-1), mask_token, patch_embeddings)
 
-        cls_token = self.cls_token.expand(batch_size, -1, -1).to(patch_embeddings.device)
-        register_tokens = self.register_tokens.expand(batch_size, -1, -1).to(patch_embeddings.device)
+        cls_token = comfy.ops.cast_to_input(self.cls_token.expand(batch_size, -1, -1), patch_embeddings)
+        register_tokens = comfy.ops.cast_to_input(self.register_tokens.expand(batch_size, -1, -1), patch_embeddings)
         embeddings = torch.cat([cls_token, register_tokens, patch_embeddings], dim=1)
         return embeddings
 

comfy/ldm/chroma_radiance/model.py

@@ -38,6 +38,8 @@ class ChromaRadianceParams(ChromaParams):
     # None means use the same dtype as the model.
     nerf_embedder_dtype: Optional[torch.dtype]
     use_x0: bool
+    # Use sequential txt_ids instead of zeros
+    use_sequential_txt_ids: bool
 
 class ChromaRadiance(Chroma):
     """
@@ -162,6 +164,9 @@ def __init__(self, image_model=None, final_layer=True, dtype=None, device=None,
         if params.use_x0:
             self.register_buffer("__x0__", torch.tensor([]))
 
+        if params.use_sequential_txt_ids:
+            self.register_buffer("__sequential__", torch.tensor([]))
+
     @property
     def _nerf_final_layer(self) -> nn.Module:
         if self.params.nerf_final_head_type == "linear":
@@ -313,6 +318,9 @@ def _forward(
         img_ids[:, :, 2] = img_ids[:, :, 2] + torch.linspace(0, w_len - 1, steps=w_len, device=x.device, dtype=x.dtype).unsqueeze(0)
         img_ids = repeat(img_ids, "h w c -> b (h w) c", b=bs)
         txt_ids = torch.zeros((bs, context.shape[1], 3), device=x.device, dtype=x.dtype)
+        # Radiance after 2026-05-22 uses sequential txt_ids instead of zeros
+        if params.use_sequential_txt_ids:
+            txt_ids[:, :, 0] = torch.arange(context.shape[1], device=x.device, dtype=x.dtype).unsqueeze(0).expand(bs, -1)
 
         img_out = self.forward_orig(
             img,

comfy/model_detection.py

@@ -313,6 +313,10 @@ def detect_unet_config(state_dict, key_prefix, metadata=None):
                     dit_config["use_x0"] = True
                 else:
                     dit_config["use_x0"] = False
+                if "{}__sequential__".format(key_prefix) in state_dict_keys: # sequential txt_ids
+                    dit_config["use_sequential_txt_ids"] = True
+                else:
+                    dit_config["use_sequential_txt_ids"] = False
         else:
             dit_config["guidance_embed"] = "{}guidance_in.in_layer.weight".format(key_prefix) in state_dict_keys
             dit_config["yak_mlp"] = '{}double_blocks.0.img_mlp.gate_proj.weight'.format(key_prefix) in state_dict_keys

comfy_api_nodes/nodes_anthropic.py

@@ -155,7 +155,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ClaudeNode",
             display_name="Anthropic Claude",
-            category="text/partner/Anthropic",
+            category="partner/text/Anthropic",
             essentials_category="Text Generation",
             description="Generate text responses with Anthropic's Claude models. "
             "Provide a text prompt and optionally one or more images for multimodal context.",

comfy_api_nodes/nodes_beeble.py

@@ -206,7 +206,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="BeebleSwitchXVideoEdit",
             display_name="Beeble SwitchX Video Edit",
-            category="video/partner/Beeble",
+            category="partner/video/Beeble",
             description=(
                 "Edit a video with Beeble SwitchX. Switches anything in the scene (background, "
                 "lighting, costume) while preserving the original subject's pixels and motion. "
@@ -302,7 +302,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="BeebleSwitchXImageEdit",
             display_name="Beeble SwitchX Image Edit",
-            category="image/partner/Beeble",
+            category="partner/image/Beeble",
             description=(
                 "Edit a single image with Beeble SwitchX. Switches anything in the scene "
                 "(background, lighting, costume) while preserving the original subject's pixels. "

comfy_api_nodes/nodes_bfl.py

@@ -37,7 +37,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="FluxProUltraImageNode",
             display_name="Flux 1.1 [pro] Ultra Image",
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Generates images using Flux Pro 1.1 Ultra via api based on prompt and resolution.",
             inputs=[
                 IO.String.Input(
@@ -155,7 +155,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id=cls.NODE_ID,
             display_name=cls.DISPLAY_NAME,
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Edits images using Flux.1 Kontext [pro] via api based on prompt and aspect ratio.",
             inputs=[
                 IO.String.Input(
@@ -277,7 +277,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="FluxProExpandNode",
             display_name="Flux.1 Expand Image",
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Outpaints image based on prompt.",
             inputs=[
                 IO.Image.Input("image"),
@@ -414,7 +414,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="FluxProFillNode",
             display_name="Flux.1 Fill Image",
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Inpaints image based on mask and prompt.",
             inputs=[
                 IO.Image.Input("image"),
@@ -521,7 +521,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="FluxEraseNode",
             display_name="Flux Erase Image",
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Removes the masked object from an image and reconstructs the background. "
             "Paint the mask over what you want to erase.",
             inputs=[
@@ -597,7 +597,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="FluxVTONode",
             display_name="Flux Virtual Try-On",
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Virtual try-on: dresses the person in the provided garment.",
             inputs=[
                 IO.Image.Input("person", tooltip="Image of the person to dress."),
@@ -697,7 +697,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id=cls.NODE_ID,
             display_name=cls.DISPLAY_NAME,
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Generates images synchronously based on prompt and resolution.",
             inputs=[
                 IO.String.Input(
@@ -868,7 +868,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="Flux2ImageNode",
             display_name="Flux.2 Image",
-            category="image/partner/BFL",
+            category="partner/image/BFL",
             description="Generate images via Flux.2 [pro] or Flux.2 [max] from a prompt and optional reference images.",
             inputs=[
                 IO.String.Input(

comfy_api_nodes/nodes_bria.py

@@ -31,7 +31,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="BriaImageEditNode",
             display_name="Bria FIBO Image Edit",
-            category="image/partner/Bria",
+            category="partner/image/Bria",
             description="Edit images using Bria latest model",
             inputs=[
                 IO.Combo.Input("model", options=["FIBO"]),
@@ -169,7 +169,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="BriaRemoveImageBackground",
             display_name="Bria Remove Image Background",
-            category="image/partner/Bria",
+            category="partner/image/Bria",
             description="Remove the background from an image using Bria RMBG 2.0.",
             inputs=[
                 IO.Image.Input("image"),
@@ -245,7 +245,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="BriaRemoveVideoBackground",
             display_name="Bria Remove Video Background",
-            category="video/partner/Bria",
+            category="partner/video/Bria",
             description="Remove the background from a video using Bria. ",
             inputs=[
                 IO.Video.Input("video"),

comfy_api_nodes/nodes_bytedance.py

@@ -368,7 +368,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceImageNode",
             display_name="ByteDance Image",
-            category="image/partner/ByteDance",
+            category="partner/image/ByteDance",
             description="Generate images using ByteDance models via api based on prompt",
             inputs=[
                 IO.Combo.Input("model", options=["seedream-3-0-t2i-250415"]),
@@ -492,7 +492,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceSeedreamNode",
             display_name="ByteDance Seedream 4.5 & 5.0",
-            category="image/partner/ByteDance",
+            category="partner/image/ByteDance",
             description="Unified text-to-image generation and precise single-sentence editing at up to 4K resolution.",
             inputs=[
                 IO.Combo.Input(
@@ -754,7 +754,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceSeedreamNodeV2",
             display_name="ByteDance Seedream 4.5 & 5.0",
-            category="image/partner/ByteDance",
+            category="partner/image/ByteDance",
             description="Unified text-to-image generation and precise single-sentence editing at up to 4K resolution.",
             inputs=[
                 IO.String.Input(
@@ -920,7 +920,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceTextToVideoNode",
             display_name="ByteDance Text to Video",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description="Generate video using ByteDance models via api based on prompt",
             inputs=[
                 IO.Combo.Input(
@@ -1048,7 +1048,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceImageToVideoNode",
             display_name="ByteDance Image to Video",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description="Generate video using ByteDance models via api based on image and prompt",
             inputs=[
                 IO.Combo.Input(
@@ -1185,7 +1185,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceFirstLastFrameNode",
             display_name="ByteDance First-Last-Frame to Video",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description="Generate video using prompt and first and last frames.",
             inputs=[
                 IO.Combo.Input(
@@ -1333,7 +1333,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceImageReferenceNode",
             display_name="ByteDance Reference Images to Video",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description="Generate video using prompt and reference images.",
             inputs=[
                 IO.Combo.Input(
@@ -1576,7 +1576,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDance2TextToVideoNode",
             display_name="ByteDance Seedance 2.0 Text to Video",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description="Generate video using Seedance 2.0 models based on a text prompt.",
             inputs=[
                 IO.DynamicCombo.Input(
@@ -1677,7 +1677,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDance2FirstLastFrameNode",
             display_name="ByteDance Seedance 2.0 First-Last-Frame to Video",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description="Generate video using Seedance 2.0 from a first frame image and optional last frame image.",
             inputs=[
                 IO.DynamicCombo.Input(
@@ -1944,7 +1944,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDance2ReferenceNode",
             display_name="ByteDance Seedance 2.0 Reference to Video",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description="Generate, edit, or extend video using Seedance 2.0 with reference images, "
             "videos, and audio. Supports multimodal reference, video editing, and video extension.",
             inputs=[
@@ -2241,7 +2241,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ByteDanceCreateImageAsset",
             display_name="ByteDance Create Image Asset",
-            category="image/partner/ByteDance",
+            category="partner/image/ByteDance",
             description=(
                 "Create a Seedance 2.0 personal image asset. Uploads the input image and "
                 "registers it in the given asset group. If group_id is empty, runs a real-person "
@@ -2308,7 +2308,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ByteDanceCreateVideoAsset",
             display_name="ByteDance Create Video Asset",
-            category="video/partner/ByteDance",
+            category="partner/video/ByteDance",
             description=(
                 "Create a Seedance 2.0 personal video asset. Uploads the input video and "
                 "registers it in the given asset group. If group_id is empty, runs a real-person "

comfy_api_nodes/nodes_bytedance_llm.py

@@ -144,7 +144,7 @@ def define_schema(cls):
         return IO.Schema(
             node_id="ByteDanceSeedNode",
             display_name="ByteDance Seed",
-            category="text/partner/ByteDance",
+            category="partner/text/ByteDance",
             essentials_category="Text Generation",
             description="Generate text responses with ByteDance's Seed 2.0 models. "
             "Provide a text prompt and optionally one or more images or videos for multimodal context.",

comfy_api_nodes/nodes_elevenlabs.py

@@ -69,7 +69,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsSpeechToText",
             display_name="ElevenLabs Speech to Text",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Transcribe audio to text. "
             "Supports automatic language detection, speaker diarization, and audio event tagging.",
             inputs=[
@@ -210,7 +210,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsVoiceSelector",
             display_name="ElevenLabs Voice Selector",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Select a predefined ElevenLabs voice for text-to-speech generation.",
             inputs=[
                 IO.Combo.Input(
@@ -239,7 +239,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsTextToSpeech",
             display_name="ElevenLabs Text to Speech",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Convert text to speech.",
             inputs=[
                 IO.Custom(ELEVENLABS_VOICE).Input(
@@ -414,7 +414,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsAudioIsolation",
             display_name="ElevenLabs Voice Isolation",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Remove background noise from audio, isolating vocals or speech.",
             inputs=[
                 IO.Audio.Input(
@@ -459,7 +459,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsTextToSoundEffects",
             display_name="ElevenLabs Text to Sound Effects",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Generate sound effects from text descriptions.",
             inputs=[
                 IO.String.Input(
@@ -555,7 +555,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsInstantVoiceClone",
             display_name="ElevenLabs Instant Voice Clone",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Create a cloned voice from audio samples. "
             "Provide 1-8 audio recordings of the voice to clone.",
             inputs=[
@@ -658,7 +658,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsSpeechToSpeech",
             display_name="ElevenLabs Speech to Speech",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Transform speech from one voice to another while preserving the original content and emotion.",
             inputs=[
                 IO.Custom(ELEVENLABS_VOICE).Input(
@@ -793,7 +793,7 @@ def define_schema(cls) -> IO.Schema:
         return IO.Schema(
             node_id="ElevenLabsTextToDialogue",
             display_name="ElevenLabs Text to Dialogue",
-            category="audio/partner/ElevenLabs",
+            category="partner/audio/ElevenLabs",
             description="Generate multi-speaker dialogue from text. Each dialogue entry has its own text and voice.",
             inputs=[
                 IO.Float.Input(