whatmodel.sh / video / Open-Sora 2.0

Open-Sora 2.0

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Open-Sora 2.0 is an open-weights video model from HPC-AI Tech, released March 2025. Its published size is 11B parameters. Licence: apache-2.0.

Model information sources

Release date ↗ · Weights licence ↗ · Parameter count ↗

11B = the Open-Sora 2.0 video diffusion model (text encoders and VAE separate).

42.4±4.7#66 of 93 video modelsfrom 2 benchmark results across 2 sources* thinly tested, treat the rank as provisional

$ how it compares

The video models ranked either side of Open-Sora 2.0, out of 139 with a score, then more from HPC-AI Tech.

#modelscorewhereRAMreleased
93Wan 2.1 I2V Turbo Alibaba43±7api price unknown2025-02
94Ray 2 Luma AI43±2comparison only2025-01
95Open-Sora 2.0 HPC-AI Tech42±5local~55GB2025-03
96Pyramid Flow (SD3) Peking University / Kuaishou42±3local~8GB2024-10
97Luma Dream Machine 1.0 Luma AI42±7comparison only2024-06
more from HPC-AI Tech
121Open-Sora 1.1 HPC-AI Tech32±7local~14GB2024-04
130Open-Sora 1.2 HPC-AI Tech27±3local~39GB2024-06

see all video models · find the ones that fit your machine

$ what it's good at

text to video▆▄▃▄▆▅█▃▂▂▂▄ #45 of 68 * · 91% of its score · 2 tests

Trails T2V-Turbo-v2, Pyramid Flow (SD3) here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
VideoVerse overall (Gemini 2.5 Pro judge)988.92 / 1608 ✎█·▄··▄·····█ beats 2/5easy · avg 1103.17 / 1608▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 53%at its level
VBench T2V total (long-prompt version, SkyReels-V2 runs)81.5% ✎█···█···█··█ beats 1/3very easy · avg 88%▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 38%typical

$ where to run it

Other ways to run it

version

  1. Run it as published · needs ~83.5 GB, NVIDIA only · from github.com/hpcaitech (official)
    git clone https://github.com/hpcaitech/Open-Sora && cd Open-Sora && pip install -v . && pip install xformers==0.0.27.post2 --index-url https://download.pytorch.org/whl/cu121 && pip install flash-attn --no-build-isolation && pip install "huggingface_hub[cli]" && huggingface-cli download hpcai-tech/Open-Sora-v2 --local-dir ./ckpts && torchrun --nproc_per_node 1 --standalone scripts/diffusion/inference.py configs/diffusion/inference/t2i2v_256px.py --save-dir samples --prompt "raining, sea" --offload True

Official weights and documentation: huggingface.co/hpcai-tech ↗

$ how the score was worked out · VideoVerse: Does Your T2V Gen… 53%, SkyReels-V2: Infinite-length… 38%

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

sourceresultsshare of the scorelicence
VideoVerse: Does Your T2V Generator Have World Model Capability to Synthesize Videos?1▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 53%CC-BY-4.0
SkyReels-V2: Infinite-length Film Generative Model1▇▇▇▇▇▇▇▇▇▇▇ 38%CC-BY-4.0
prior · pull toward the middle▇▇▇ 9%
weighted average= 42.4

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

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