MiniMax Text-to-Video (fal)

MiniMaxTextToVideo is the no-frills text-to-video node: one prompt, one video. It calls MiniMax's video model (also branded Hailuo) - one of the stronger closed video generators - through fal, with no starting image and no pile of settings. Type what you want to see, get a clip. It runs on fal's servers via gokayfem's ComfyUI-fal-API pack, so nothing's local and you pay per generation.

Use it when you're starting from words, not a picture, and you don't want to think about knobs. For animating an existing image, use MiniMax (image-to-video) in the same pack; for keeping a specific subject consistent, MiniMaxSubjectReference.

How it works

The node sends your prompt to fal, MiniMax generates the video on its servers, and you get back a URL to the clip. Everything else - resolution, length, motion defaults - is decided server-side. As with the other video nodes, the output is a string URL, so download or save it before the link expires.

The inputs and outputs that matter

• prompt - the only input, and it's carrying the whole generation. Describe the scene and the motion, because there's no image to establish either. "A drone shot flying low over a misty pine forest at dawn, slow forward motion" gives it far more to work with than "a forest."

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.
• Bland or generic results. With only a prompt to go on, detail is everything - describe camera, subject, and motion. A vague prompt gets a vague video.
• 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 actual error - content policy, quota, or a bad input. Read it.
• Cost. Every 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.

It's the simplest way to get a MiniMax clip from a blank page. Put your effort into a detailed, motion-aware prompt and this node does the rest - and when you want more control, the image-to-video and subject-reference siblings are one node away.