#6363rd of 89 in Image and videoFor image, video and design, video and animation
mochi
10B-parameter text-to-video diffusion model with inference code and LoRA trainer
Live demo ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 3.8k
- License
- Apache-2.0
- Last commit
- Oct 2026
- Language
- Python
3,613 on 25 Apr 20263,768 on 14 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Mochi 1 is Genmo's open-weight text-to-video model, a 10B AsymmDiT diffusion transformer paired with a 362M AsymmVAE and a T5-XXL text encoder. The repo provides a Gradio UI, a CLI, a Python pipeline API, and a LoRA fine-tuning trainer. Output is 480p, and the native implementation needs about 60GB VRAM on a single GPU.
Who it is for: Engineers running or fine-tuning open text-to-video models on high-end GPUs
Strengths
- Apache-2.0 weights, with the 10B DiT and VAE both released
- Ships Gradio UI, CLI and a composable Python pipeline API
- LoRA fine-tuning on a single H100 or A100 80GB
- Supports multi-GPU context parallel inference as well as single GPU
Weaknesses
- Needs about 60GB VRAM on one GPU; H100 recommended
- Generates 480p only, with warping on extreme motion
- Weak on animated content; tuned for photorealistic styles
- README says GPU inference and fine-tuning still need qualification before deployment
What it needs
- GPU required
- Needs FFmpeg, uv, Python 3.11, T5-XXL text encoder
- Models: Mochi 1 preview (AsymmDiT 10B, safetensors), AsymmVAE, LoRA safetensors
Also in Image and video
See all 89| Rank | Project | Score |
|---|---|---|
| 1 | frigateLocal NVR with realtime object detection for IP cameras | 93 out of 100 |
| 2 | MoneyPrinterTurboGenerates short videos from a topic, with script, footage, voiceover and subtitles | 88 out of 100 |
| 3 | toonflow-appInfinite-canvas tool for AI short drama and comic video production | 88 out of 100 |
| #4 | InvokeAIWeb UI for diffusion image generation with inpainting canvas and node workflows | 87 out of 100 |
| #5 | ComfyUINode-graph editor for running diffusion image, video, audio and 3D workflows | 85 out of 100 |