Mochi 1
Best for: Open AI video research
Some links are affiliate links; we may earn a commission at no extra cost to you.
Genmo's open text-to-video model for research, experimentation and local video generation.
Key features
- ✓Text-to-video
- ✓Open weights
- ✓Local inference
- ✓Research
- ✓Developer experimentation
Overview
Mochi 1 is an open video generation model developed by Genmo. It is designed for text-to-video generation and is particularly useful for researchers and developers experimenting with open generative video.
How Mochi 1 works
Mochi 1 is an open text-to-video model from Genmo. Its weights are available for research and experimentation, so you can run it locally on suitable hardware and inspect how it behaves.
It is aimed at people who want to learn from or build on an open model, rather than a polished consumer app. Hosted options exist for those who prefer an API on a usage basis.
Getting started with Mochi 1
- Read the project page for the current release, licence and GPU requirements.
- Set up the model locally or pick a hosted endpoint.
- Generate a short clip with a simple prompt to establish a baseline.
- Experiment with prompt style and settings, noting results so you can compare changes.
Is Mochi 1 the right choice?
Pick Mochi 1 for open research and experimentation. If you need dependable commercial-grade output with minimal setup, a hosted closed model is simpler.
Who should use Mochi 1
Researchers, developers and AI enthusiasts interested in open video-generation models and local experimentation.
Try Mochi 1
Ready to try it? Visit the Mochi 1 official website to review current access, usage limits and pricing, then run a few test prompts of your own before committing to a plan.
Pros & cons
- +Open model
- +Research friendly
- +Local experimentation
- +Developer access
- −Hardware requirements
- −Older than newer commercial models
- −Technical setup required
Pricing
Open Model
- ✓Open weights
- ✓Local inference
Hosted
- ✓Hosted inference
What does Mochi 1 cost for longer videos?
Usage-based pricing is charged per second of generated video, so cost grows with length. Costs below are a single clean pass at list price (reviewed October 2026). Most clips need 2 to 4 attempts, so multiply by your expected retakes for a realistic budget.
| Model · route | Per second | 10 sec | 30 sec | 1 min | 5 min | 30 min |
|---|---|---|---|---|---|---|
| Mochi 1Self-hosted GPU · 480p · estimate | ~$0.042 | $0.42 | $1.25 | $2.50 | $12.50 | $75.00 |
- Mochi 1: Open weights. Generates at 480p; estimate of GPU minutes per second of video.
Specifications
| Developer | Genmo |
|---|---|
| Text-to-Video | Yes |
| Open Model | Yes |
| Local Inference | Yes |
| Developer Use | Yes |