MiniMax H3 ComfyUI Workflow: Make Long Videos With OBVPM Timeline @ComfyUIworkflows
MiniMax H3 ComfyUI Workflow: Make Long Videos With OBVPM Timeline  @ComfyUIworkflows
Uploaded October 2026 | Updated November 2026, 1 day ago
Make long MiniMax H3 videos in ComfyUI, where the character and the place stay the same from clip to clip. I extend the video one clip at a time with the OBVPM Timeline custom node, then upscale the whole video at the end.

I tried the MiniMax H3 extend nodes I could find, and this is the one I use right now. It saves a small .mctx file next to each clip. That file keeps the clip's latent, so the next clip continues from that data, not only from the last frame.

In this video:
- Two examples: a fight scene made with reference images only (no background image), and a running scene made with a background image and a reference image
- How to make the first clip, and why I keep upscale off while testing
- How to accept or reject each new clip
- How to extend a clip, and what the window settings 39, 90 and 141 mean (they are frames, not seconds)
- How to cut a bad part of a clip and continue from the good part
- How I added a new character, a ghost, to the scene
- My test: window 39 vs window 90. For this scene, 39 worked better
- How to fix an older clip with Load Settings, without starting the whole video again
- How to upscale the whole timeline in one pass

It is not perfect. Some clips go wrong, and I show how I fixed them.

Three tips from the OBVPM developer:
1. Do not change the resolution after your first clip.
2. Keep each video file and its .mctx file together. Do not edit or re-save that video. Edit a copy instead.
3. If the quality drops after many extensions, generate a fresh clip and join it with Bridge.

Resources
Workflow: patreon.com/ComfyuiWorkflowBlog/posts/minimax-h3-long-171392521
OBVPM Timeline (GitHub): github.com/chanon/comfyui-obvpm-timeline
OBVPM companion nodes (GitHub): github.com/chanon/comfyui-obvpm
Check the GitHub page for the ComfyUI and node versions the workflow needs.

Which window will you try first, 39 or 90? Tell me in the comments.

#MiniMaxH3 #ComfyUI #AIVideo
MiniMax H3 ComfyUI Workflow: Make Long Videos With OBVPM TimelineSora 2 in ComfyUI: Step-by-Step Free Workflow to Generate 15-Sec AI Video NewKrea 2 LoRA Training: 1-Click Local Setup & Cheapest Cloud WayComfyUI Text-to-Video: Create Videos Like Runway Gen-3 with 8GB VRAM Text to Realistic #comfyuiWAN 2.2 ComfyUI Workflow: Infinite, Non-Looping Video on Low VRAM (Q5 GGUF)Hunyuan AI Text to Video in Seconds: Low VRAM | Hunyuan Video 6-Step ComfyUI WorkflowKrea 2 ComfyUI: Style Transfer & Perfect Text (Free Workflow)MiniMax H3 3-Step LoRA: Is It Really 11x Faster?ComfyUI Flux: How to Create Your Own FLUX LoRAs Using Flux Gym: Low VRAMQwen Image 2.1 Finally Has ControlNet: Pose, Depth and Inpainting TestedQwen Image Low-VRAM: Text & Img2Img + Realism in ComfyUI (Q4 GGUF vs BF16)Qwen-Image vs Flux Krea: ComfyUI Workflow with GGUF for Text-to-Image — Here’s What Worked
ComfyUI Workflow Blog |

MiniMax H3 ComfyUI Workflow: Make Long Videos With OBVPM Timeline

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER