whatmodel.sh / video / Mochi 1
Mochi 1
↗ share
+ contribute
Mochi 1 is an open-weights video model from Genmo, released October 2024. Its published size is 10B parameters. Licence: apache-2.0.
$ how it compares
The video models ranked either side of Mochi 1, out of 139 with a score.
| # | model | score | where |
|---|---|---|---|
| 98 |  Sora (original, 2024) | 42±2 | comparison only |
| 99 |  Wan 2.1 T2V 1.3B | 40±4 | api $0.040/sec +local |
| 100 | Mochi 1 | 40±2 | api price unknown +local |
| 101 |  Runway Gen-3 Alpha | 40±3 | comparison only |
| 102 |  Open-Sora Plan v1.3 (I2V) | 39±7 | local |
$ what it's good at
text to video▆▄▃▄▆▅█▃▂▂▂▄ #48 of 68 · 88% of its score · 4 tests
Trails LTX-2.3, T2V-Turbo-v2 here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LMArena video | 1006.6 ✎ | ▃▂▃▆█▆·▆▅▁▂▇ beats 0/41 | 49% |
| GenAI-Arena text-to-video (anonymous Elo) | 1121.3 ✎ | ▃▃···▅▃·▃▅·█ beats 9/10 | 26% |
| VideoPhy (auto-evaluator average) | 41.0% ✎ | ▅·▃▅··▃█···▃ beats 8/9 | 7.0% |
| T2VPhysBench (human physical consistency) | 34.0% ✎ | █··█·█·····█ beats 2/3 | 5.7% |
other tests2 not in a task · 11% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Wan-Bench (Wan2.1 release) weighted score | 63.9% ✎ | █·······█··█ beats 0/2 | 6.0% |
| WorldModelBench (human total) | 7.62 / 10 ✎ | ▅··▅▃···▃··█ beats 7/8 | 5.0% |
$ where to run it
Through an API
api · Replicate · price unknown · replicate.com ↗Price unknown. Public model run/API page. Compute-time pricing, where applicable, is not converted to output-unit pricing.
Other ways to run it
version
- Run it as published · needs ~24.5 GB, NVIDIA only · from huggingface.co/genmo (official)
python -c "import torch;from diffusers import MochiPipeline;from diffusers.utils import export_to_video;p=MochiPipeline.from_pretrained('genmo/mochi-1-preview',variant='bf16',torch_dtype=torch.bfloat16);p.enable_model_cpu_offload();p.enable_vae_tiling();export_to_video(p('Close-up of a chameleon eye, its scaly skin changing color',num_frames=84).frames[0],'mochi.mp4',fps=30)"
Official weights and documentation: huggingface.co/genmo ↗
$ how the score was worked out · LMArena leaderboard dataset 49%, GenAI-Arena (TIGER-Lab) text-… 26%, VideoPhy / VideoPhy-2 officia… 7% and 3 more
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 LMArena leaderboard dataset | 1 | 49% |
 GenAI-Arena (TIGER-Lab) text-to-video leaderboard | 1 | 26% |
 VideoPhy / VideoPhy-2 official repository leaderboards | 1 | 7% |
| Wan-AI/Wan2.1-T2V-14B model card | 1 | 6% |
 T2VPhysBench: A First-Principles Benchmark for Physical Consistency in Text-to-Video Generation | 1 | 6% |
| WorldModelBench: Judging Video Generation Models As World Models | 1 | 5% |
| prior · pull toward the middle | 1% | |
| weighted average | = 39.9 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json