Is this the final form of open-source Digital Humans? 🤖
Today, we push ComfyUI to its limits with the new "Infinite Talk" nodes! I'm sharing 7 distinct workflows that let you generate UNLIMITED duration videos with perfect lip-sync, using Wan2.1 and even the newer Wan2.2 for superior motion control.
Whether you want a Single Speaker, a Dual Dialogue, or a Face-Swap Video, this tutorial covers it all.
- - - - - - - - - - - - - - - - - - - - - - - - - -
🔽 GET THE FREE WORKFLOWS & 1000 BONUS COINS 🔽
✨ Register on RunningHub through my link to get 1000 FREE RH Coins!
💰 Plus, claim 100 extra coins every single day just by logging in.
1️⃣AI APP: InfiniteTakl(Single & Unlimited duration)Cycle Version:
https://www.runninghub.ai/ai-detail/2020107483410337794/?inviteCode=rh-v1291
2️⃣AI APP: InfiniteTakl(Single & Unlimited duration)Original version:
https://www.runninghub.ai/ai-detail/2020195096758587393/?inviteCode=rh-v1291
3️⃣AI APP: InfiniteTakl(Two-person& Unlimited duration)
https://www.runninghub.ai/ai-detail/2020446660488335362/?inviteCode=rh-v1291
4️⃣AI APP: InfiniteTakl(Two_person & Unlimited duration)Original version:
https://www.runninghub.ai/ai-detail/2020448621434507266/?inviteCode=rh-v1291
5️⃣AI APP: wan2.2+InfiniteTakl(Unlimited duration):
https://www.runninghub.ai/ai-detail/2019999909776199682/?inviteCode=rh-v1291
6️⃣AI APP: wan2.2+InfiniteTalk(First and last frames):
https://www.runninghub.ai/ai-detail/2020047059822321666/?inviteCode=rh-v1291
WORKFLOWS:
1️⃣Workflow: InfiniteTakl(Single & Unlimited duration)Cycle Version:
https://www.runninghub.ai/post/2020048650625355778/?inviteCode=rh-v1291
2️⃣Workflow: InfiniteTakl(Single & Unlimited duration)Original version:
https://www.runninghub.ai/post/2020151568548499457/?inviteCode=rh-v1291
3️⃣Workflow: InfiniteTakl(Two-person& Unlimited duration):
https://www.runninghub.ai/post/2020104313544122370/?inviteCode=rh-v1291
4️⃣Workflow: InfiniteTakl(Two_person & Unlimited duration)Original version:
https://www.runninghub.ai/post/2020192003224506370/?inviteCode=rh-v1291
5️⃣Workflow: wan2.2+InfiniteTakl(Unlimited duration):
https://www.runninghub.ai/post/2019682157827989506/?inviteCode=rh-v1291
6️⃣Workflow: wan2.2+InfiniteTalk(First and last frames):
https://www.runninghub.ai/post/2020007962508529666/?inviteCode=rh-v1291
7️⃣Workflow: wan2.1+InfiniteTalk(V2V_FaceSwap):
https://www.runninghub.ai/post/2019839305782403073/?inviteCode=rh-v1291
Models & Plugin:
1️⃣ComfyUI-PainterAI2V
https://github.com/princepainter/ComfyUI-PainterAI2V
2️⃣Comfyui-PainterAudioCut
https://github.com/princepainter/Comfyui-PainterAudioCut
Model Collection:
Download link for InfiniteTalk (Online storage):
https://pan.quark.cn/s/79879ae18e50
- - - - - - - - - - - - - - - - - - - - - - - - - -
🚀 What's Inside This Tutorial:*
* Infinite Loop Logic: How to calculate frames for endless video generation.
* Dual Speaker Mode: Creating conversations between two AI characters.
* Wan2.2 Integration: Using the latest model for better body movement and camera control.
* Video Face Swap: Keeping the original video motion but changing the face and lip-sync.
If this video helps you master AI Video, please SMASH that LIKE button and SUBSCRIBE! Let's build the future of AI together.
#ComfyUI #InfiniteTalk #Wan2 #Wan2_1 #AIVideo #TextToVideo #DigitalHuman #StableDiffusion #TalkingAvatar #RunningHub
Description
FAQ
Comments (9)
wow, this is phenomenal, just watched your tutorial on youtube and subscribed immediately.
I will definitely test this off hours, I wonder if it works with Cache-DIT and SVI pro two loras
🤔 I'm not sure, but if you try it and it works, could you come back and let me know? 😆
PainterAI2V node is missing
To run locally, you'll need to download this node:https://github.com/princepainter/ComfyUI-PainterAI2V
did that, but still its not working, showing like missing node
@myprivacy27091991221 It might be an issue with your environment configuration. Could you confirm if your ComfyUI core is updated to the latest version? If everything is already up to date, I'll see if I can provide feedback to the developer. Alternatively, you could try using an online platform for now.
Dude, update your ComfyUI, or install from manager https://github.com/princepainter/ComfyUI-PainterAI2V.git
Excuse me,Can 12GB VRAM run it?
You can try lowering the output resolution, increasing the system's virtual memory, or switching to the GGUF version of the model.