ComfyUI: Running MultiTalk on RunPod
ComfyUI
Demo summary
The creator walks through executing the MultiTalk high-quality workflow on a RunPod instance, monitoring VRAM usage with nvitop while generating the video.
Step-by-step
- Drag and drop the 720p long context high quality workflow into ComfyUI
- Upload your demo image and audio file
- Set the width and height to match your image (e.g., 720x1280)
- Calculate and enter the number of frames based on audio length (25 frames per second)
- Enter the text prompt describing the character's action
- Set 'blocks to swap' to 0 for high VRAM machines or 40 for low VRAM machines
- Click the Run button at the bottom of the interface
- Monitor the generation progress via the terminal or nvitop
- Right-click the result and select Save Preview to download the video
Options
- Enable 'tiled VAE' or 'all tiling' if you encounter out-of-memory errors
- Bypass the block swapping node on high VRAM machines to speed up model loading
- Download the entire output folder as an archive from the workspace file browser
Watch out for
- The number of frames is not automatically set; you must manually calculate it based on audio duration
- Initial model loading on RunPod can take a significant amount of time
- High-quality processing requires 10 steps and significant VRAM (up to 35GB for 720p)
Tips
- Use 25 frames for every one second of audio
- Monitor nvitop to verify the GPU is being fully utilized (e.g., checking watt usage and VRAM)
- Check the terminal status to see the exact duration of the uploaded audio for more accurate frame calculation
Highlights
“the machine is ready and set... the rest of the usage is exactly same as in the Windows tutorial part.”
All demos from “MultiTalk Full Tutorial With 1-Click Installer - Make Talking and Singing Videos From Static Images”
- 7:406:12Generate 480p talking video in ComfyUIThe creator demonstrates loading a static image and audio file into a ComfyUI workflow using the MultiTalk node to generate a 10-second talking animation.ComfyUI· AI Animation Generator
- 13:524:06High-quality 720p long context generationA demonstration of the 720p long context workflow in ComfyUI, showing how to adjust resolution, prompt, and block swap parameters for higher fidelity output.ComfyUI· AI Animation Generator
- 18:560:49Side-by-side video quality comparisonThe creator uses an 'Ultimate Video Upscaler' tool to perform a side-by-side comparison between the 480p and 720p generated outputs.ComfyUI· AI Video Upscaler
- 47:445:08Running MultiTalk on RunPodCurrentThe creator walks through executing the MultiTalk high-quality workflow on a RunPod instance, monitoring VRAM usage with nvitop while generating the video.ComfyUI· AI Animation Generator
- Watch “MultiTalk Full Tutorial With 1-Click Installer - Make Talking and Singing Videos From Static Images” →
AI Animation Generator
1:420:59Seed hunting with a multi-stage LTX 2.3 workflowThe creator demonstrates his custom ComfyUI workflow that generates four low-resolution LTX 2.3 samples simultaneously to find a 'golden seed' before upscaling to 1080p.Fox•Fur•Essence Films
16:361:56Setting up HunyuanVideo 1.5 in ComfyUIThe video demonstrates how to update ComfyUI and import the HunyuanVideo 1.5 JSON workflow files to create a node-based generation environment.AI Search- 20:151:53Text-to-Video generation in ComfyUIA step-by-step demo of configuring the Hunyuan nodes in ComfyUI, entering a prompt for a 'giant cat', and rendering the final 720p video.AI Search
- 29:151:33Running HunyuanVideo with GGUF (Low VRAM)The video shows how to use the GGUF loader node to run a compressed version of HunyuanVideo 1.5, enabling video generation on GPUs with as little as 6GB of VRAM.AI Search
0:510:36Configure Infinite Talk and Wan 2.1 models in ComfyUIThe user demonstrates loading the Infinite Talk model alongside the Wan 2.1 I2V 14B model within ComfyUI, including enabling block swap and torch compile for VRAM optimization.Olares- 1:551:13Configure sampling and window settings for long video generationThe user walks through the Wan Video Wrapper sampling node, explaining how to set frame window size, motion frame overlap, and start steps for consistent video generation.Olares
0:310:46Generate cinematic video with LTX MSR workflowThe creator demonstrates using the LTX MSR workflow in ComfyUI to generate a video from multiple reference images and a prompt, highlighting the 3D camera movement and character consistency.Apex Artist
0:290:24Load source footage and models in ComfyUIThe user demonstrates importing source video footage and loading the necessary model nodes including WAN Video, VAE, and Clip Vision within the ComfyUI interface.ComfyUI
0:431:12Configure LTX-2.3 MSR LoRA and Prompt Relay in ComfyUIThe creator demonstrates setting up a ComfyUI workflow using the LTX-2.3 model with the MSR LoRA and Prompt Relay nodes to manage video generation on 8GB of VRAM.bigboss97
4:261:28Overview of the TensNodes Consistent Character Workflow in ComfyUIThe creator walks through a four-stage ComfyUI workflow utilizing TensNodes to process reference images and text for stable LTX video generation.SOTAI
3:411:25Configure LTX 2.3 foundational parametersA walkthrough of setting dimensions, frame counts, and loading core models including the LTX 2.3 distill model and audio VAE within a ComfyUI workflow.SOTAI- 5:060:37Process visual and audio inputs in ComfyUIThe demo shows how to use the LTX V image-to-video condition node and empty latent audio node to prepare data for the generation engine.SOTAI