Comfy Agent
▶ Play demoYouTube · loads on clickWhat they did
ComfyUI's built-in agent plans, wires and debugs workflows on your canvas. Describe what you want to make and it connects the nodes and models, then fixes what breaks.
Worth knowing
- Price: unknown
- Status: sourced. We have not run this ourselves yet (no "Tested" badge).
- Quality bar (computed by code): 5/5 (primary source ✓, reproducible ✓, new ✓, example output ✓, safe ✓)
The best new AI video workflows and prompts, weekly.
Related
▶ demo at sourceComfyUI
ComfyUI is the open, node-based app for building image and video pipelines on your own GPU or in the cloud. Most open-model video workflows assume it.
▶ demo at sourceComfyUI docs: video
Official ComfyUI tutorial for Wan 2.1: text-to-video with the 1.3B model and image-to-video with the 14B model, including where each model file goes.
▶ demo on XComfyUI Changelog
ComfyUI's release log. Check it to see when new video models get native nodes.
▶ YouTube demoBlender blockout as the blueprint for a ComfyUI video shot
ComfyUI shows a 13-second anime fight where a simple Blender blockout fixes camera movement, character placement, attack paths and timing, while reference images supply the final characters and style. Structure first, style second. The description links a runnable workflow.
▶ YouTube demoKeyframes to full character animation in ComfyUI (Wan 2.2 FLF)
The official ComfyUI tutorial for in-betweening: you draw key poses, each consecutive pair runs through Wan 2.2's first-last-frame model, and the clips are thinned, assembled in order and rendered as one animation.
▶ YouTube demoLow-VRAM character swap with MiniMax H3 in ComfyUI (no masking)
A ComfyUI workflow that swaps a character in a video using MiniMax H3 plus a Character Swap LoRA, with no masking. It renders at about 0.4–0.5 MP in 8 steps and then upscales to 1.2 MP with the H3 latent upscaler. The creator tested it on an RTX 3060 6 GB with 16 GB RAM: about 20 minutes to generate and 21 to upscale. You build a character sheet first (Qwen Image 2.1, FLUX Krea or H3). Workflow and tutorial are linked from the post.