Genmo

What is Genmo?

Genmo is an AI lab that builds video generation models and describes its goal as world models that understand the physical world. Its main release is Mochi 1, an open-source text-to-video model available through a web playground, GitHub, ComfyUI and Hugging Face.

Key capabilities of Genmo

• Text-to-video generation: turns written prompts into short video clips with Mochi 1
• Open weights: Mochi 1 can be downloaded from GitHub or Hugging Face
• Local use: runs on your own hardware via the repository or ComfyUI
• Web playground: browser interface to test prompts without setup
• Replay feature: a lower-cost generation option billed at 50 credits
• Commercial use: paid tiers remove the watermark and allow commercial usage

How Genmo works

You write a text prompt in the playground and the Mochi 1 model renders a video, which costs 100 credits per generation. Developers can instead download the model and run it locally. Monthly credits reset each billing cycle, and paid plans add queue priority.

Who uses Genmo?

Creators and developers who want an open video model they can test online or self-host, and researchers who build on the released weights.

Genmo pricing

The Free plan includes 250 one-time credits, unlocked by adding a payment method, and videos carry a Genmo watermark. Lite gives 1,200 monthly credits and Standard gives 5,000, both without watermark and with commercial rights. Lite at $10 and Standard at $30 per month come from third-party listings, and yearly billing is cheaper.

Genmo alternatives

Hedra and Arcads focus on avatar and ad video, Moonvalley trains Marey on licensed footage, and Luma Dream Machine and Sora are hosted text-to-video generators.