Genmo (Mochi)

Genmo builds the open-source Mochi video models and a creative platform for generating expressive AI video from text and images.

Overview

Genmo is the company behind Mochi 1, an open-source text-to-video model that surprised the industry with its motion quality and prompt adherence when released. Weights are published for the community to run and fine-tune, continuing the open-model tradition Stable Diffusion started for images. The hosted Genmo platform adds a polished creation experience with text-to-video, image-to-video, and directable camera motion.

Mochi’s later versions push resolution and duration further, and Genmo’s product work emphasizes expressiveness — smooth movement, dynamic shots, and playful stylization — rather than photorealism alone. For developers and researchers, the open weights mean the model can be studied, adapted, and deployed on custom infrastructure.

Open weights make Genmo unusually transparent: researchers can inspect the architecture, reproduce results, and fine-tune on domain footage — for scientific visualization, game prototyping, or archival animation. The hosted product keeps improving alongside the open releases, giving casual users the benefits of the same research.

Key Features

• Mochi 1 open-source text-to-video model with strong motion quality
• Hosted Genmo platform for text-to-video and image-to-video generation
• Directable camera motion and stylization controls
• Open weights available for self-hosting and fine-tuning
• Research-forward roadmap with rapid model improvements
• Creator community sharing prompts and results
• Free generation tier with paid speed and quality upgrades
• Self-hosting and domain fine-tuning enabled by open weights
• Hosted product improvements that track the open research roadmap

Pricing

Genmo offers a free tier for casual generation, with paid plans unlocking faster queues, higher resolutions, and commercial usage rights. The open-source Mochi weights are free under their published license for self-hosting. Paid plans are modest relative to closed competitors, reflecting Genmo’s goal of keeping expressive video generation broadly accessible.

Who Is It For

Genmo serves AI video enthusiasts, researchers, and developers who want open models plus a friendly hosted experience.

It fits creators who value expressive, stylized motion over strict photorealism.

Startups building video features often prototype on Mochi’s open weights before deciding whether to pay for hosted inference at scale.

FAQ

Is Genmo free to use?

There is a free tier for hosted generation, and the Mochi model weights are freely downloadable under an open license.

Does Genmo offer an API?

Yes — hosted generation is available through Genmo’s platform, and the open model can be self-served through community tooling.

How does Genmo compare to Runway or Pika?

Genmo stands out through its open-source models and expressive motion, while Runway offers pro editing depth and Pika emphasizes quick, playful clip creation.

Can I run Mochi models locally?

Yes — Mochi model weights are published openly, so you can run inference on suitable GPU hardware, fine-tune on your own data, and deploy without licensing fees beyond the model’s published terms.