Where ComfyPods is going — gathering the place where video happens at scmf.ai
- The service
- scmf.ai
- Japanese entry
- aicu.jp/comfy
- Ways to get it
- 3
- Price / date
- not decided
In three lines
The service itself converges on scmf.ai. The Japanese entry stays at aicu.jp/comfy. We are making the environment where MiniMax H3 video is possible actually usable — that is what we most want to do right now. Three ways to get it, and no casual subscription.
Being able to make video is itself the value
For a month we have been making video on ComfyUI and writing it up. Five seconds going from 16 minutes to 3.8. Vertical framing making the model paint a scene instead of a green screen. Six characters looping. What we learned is that video does not start without somewhere to try it. Get a GPU, gather the models, line up the versions, and only then does the first clip appear. Most people stop before that. ComfyPods carries that part for you.
Three ways to get it. No subscription
We decided not to build a "just subscribe" plan. What matters is that someone who wants to make video can make it when they want to. (1) Pay for what you use, when you use it — a one-time payment, not a subscription. (2) Invitations and coupons, for former AICU Lab+ users, including a free period from an invitation. (3) Enterprise enquiries, for data that cannot leave your organisation. Pricing is not decided. Writing down what is not decided turns into a broken promise later. We will announce it here when it is.
The version holds
ComfyUI ships almost weekly. That is good, but it means what worked yesterday may not work today. The version ComfyPods runs is pinned with ComfyLTS — the core and every custom node pinned by SHA. An upstream release does not change your environment that day. We check new versions on Nightly first, and only take them in once we have seen an image come out.
What is not decided
Honestly: pricing for each of the three. When it becomes available — we will not promise a date here. We previously wrote "planned for September 2026" and did not make it. And how migration from the existing shared ComfyUI will work. Not writing down what is not decided is the part we care about most.
A note on the MiniMax H3 licence
Why this page carries no MiniMax H3 output (2026-09-15). The MiniMax H3 Community License limits the licence to an Applicable Territory — "worldwide, excluding the Excluded Territories" — and names those as the European Union, the United Kingdom, the Republic of Korea and the United States of America (§I.3, §I.5). §III.4 forbids using, reproducing, distributing or displaying the Works or any of their Outputs outside that territory. This English edition is read from those places, so we removed the H3 clips and stills from it. The measurements stay — they are ours, not outputs of the model. Licence text: https://huggingface.co/MiniMaxAI/MiniMax-H3/raw/main/LICENSE
Related
- Is FLUX really non-commercial? We counted 312 car and train LoRAs
- MiniMax H3 ref2va with the 4-step Turbo LoRA: 16.1 min to 3.8 min at 1344x768
ComfyPods runs ComfyUI in your browser. No install, no GPU of your own. Open beta; video generation is planned for September 2026.
→ Newsletter for release notices
→ ComfyLTS — our long-term-support ComfyUI package. A star tells us how many people need this.