MiniMax Video Generation (fal)

MiniMax's video model (you'll also see it branded Hailuo) is one of the stronger closed video generators - the kind that consistently lands near the top of comparison threads for motion and prompt-following. MiniMax is the image-to-video node for it: give it a still image and a prompt, and it animates the picture. It runs on fal's servers through gokayfem's ComfyUI-fal-API pack, so there's no local model and no VRAM - just a per-clip bill.

This is the node when you already have the frame you want and you want it to move. Great for bringing a generated image or a photo to life without describing the whole scene from nothing.

How it works

Both inputs are required, which tells you what it's for: an image plus a prompt, animated. fal runs the generation remotely and returns a URL to the resulting clip. As with the other video nodes here, the output is a string URL - download or save it before the link expires.

The inputs and outputs that matter

• image (required, IMAGE) - the frame to animate. This is an image-to-video node; there's no text-only path.
• prompt (required) - describe the motion and what should happen. Since the image already establishes the scene, spend your words on movement: "slow push-in, hair moving in the wind," not a full scene description.

That's the whole surface - deliberately minimal. The single output is a STRING: the video URL.

How to install it

Via ComfyUI Manager: search ComfyUI-fal-API, install, restart. Manually:

cd ComfyUI/custom_nodes
git clone https://github.com/gokayfem/ComfyUI-fal-API.git
pip install -r ComfyUI-fal-API/requirements.txt

No models download. Add a fal API key to config.ini under [API] (from config.ini.example) or export FAL_KEY=your_key, restart, and find it under FAL/VideoGeneration.

Common issues & troubleshooting

• Auth error. FAL_KEY not set. Fix and restart.
• Nothing to run. Both image and prompt are required - connect an image source and write a prompt, or the node has nothing to do.
• The motion isn't what you asked for. Describe movement specifically. The image fixes the content, so the prompt's job is direction and action, not scenery.
• The URL expired. fal result links have a shelf life - route the STRING output into the pack's Save-Media-from-URL utility or download it promptly.
• A run fails. The pack surfaces fal's real error text - content policy, quota, or a bad input. Read it.
• Cost. Each clip is a paid API call. The result cache serves repeat identical calls free (force_rerun bypasses); the spend guard caps a session.
• Windows Portable. Missing fal-client? python_embeded\python.exe -m pip install fal-client.

If you want a different flavor of image-to-video than Kling or Luma, MiniMax is well worth a spin - feed it a strong starting frame and a motion-focused prompt and it tends to deliver. For pure text-to-video with no image, use the MiniMaxTextToVideo node in the same pack instead.