[For Patreon Supporters] Wan I2V - InfiniteTalk - Long Content Talking Avatar Generator (Single) (Ver.20250822)
Added 2025-08-22 13:00:18 +0000 UTC
Related Tutorial : https://www.patreon.com/posts/137034879
(Check it out if you want to learn how to config)
MeiGen-AI/InfiniteTalk
https://github.com/MeiGen-AI/InfiniteTalk
InfiniteTalk for ComfyUI
https://huggingface.co/MeiGen-AI/InfiniteTalk/tree/main/comfyui
ComfyUI_ChatterBox_SRT_Voice
https://github.com/diodiogod/ComfyUI_ChatterBox_SRT_Voice
For ComfyUI_ChatterBox_SRT_Voice Setup:
cd ComfyUI/custom_nodes
git clone https://github.com/diodiogod/ComfyUI_ChatterBox_SRT_Voice.git
cd ComfyUI_ChatterBox_SRT_Voice
pip install -r ComfyUI_ChatterBox_SRT_Voice/requirements.txt
Download Models (Required)
Download from HuggingFace ChatterBox (https://huggingface.co/ResembleAI/chatterbox/tree/main)
Place in ComfyUI/models/chatterbox/
Required files:
conds.pt (105 KB)
s3gen.pt (~1 GB)
t3_cfg.pt (~1 GB)
tokenizer.json (25 KB)
ve.pt (5.5 MB)
Comments
1 - try other number for the Model blocks swap 2 - llm_response Set node ? it's there Bypass in default. have you check out the tutorial how it run? or maybe make sure you choose the right text input for TTS ?
Benjamin Law
2025-09-04 18:46:28 +0000 UTCHello. I can't run the workflow; it gives the error: WanVideoTextEncode 'blocks.1.pos_embedding.embedding.weight' and also No SetNode found for llm_response(GetNode). Most likely, you're missing custom nodes.
Ivan Carlos Horbach
2025-09-04 17:00:17 +0000 UTCHello Benji. The tts srt isn't available in runpod to install directly from comfy manager. i cloned the repo from github and is not the same node. what should i do ? i updated comfy 3.52. Thanks. Edit: Solved ! apt-get update apt-get install -y portaudio19-dev libportaudio2 libportaudiocpp0 pip install --force-reinstall sounddevice PyAudio for those who have the same problem !
Micu George
2025-08-24 22:00:29 +0000 UTCWhats your setting?
Benjamin Law
2025-08-23 07:03:38 +0000 UTChey mate, this isnt working for me at all. Cant get the reference image to show up, getting some fuzzy random person. My audio source is 26 seconds and the generation is going for 2s every generation. Can you please advise or send a fix. Thanks
Scott Reactz
2025-08-23 06:45:35 +0000 UTC
