whatmodel.sh / video / Allegro

Allegro

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

Allegro is an open-weights video model from Rhymes AI, released October 2024. Its published size is 2.8B parameters. Licence: apache-2.0.

35.0±2.3#79 of 93 video modelsfrom 3 benchmark results across 2 sources

$ how it compares

The video models ranked either side of Allegro, out of 139 with a score.

#modelscorewhereRAMreleased
114CogVideoX-2B Zhipu AI (THUDM)36±2local~5.3GB2024-08
115Pika 2.2 Pika36±2api $0.040/sec2025-02
116Allegro Rhymes AI35±2local~9.3GB2024-10
117Kling 1.0 Kuaishou (Kling AI)35±7comparison only2024-06
118VideoCrafter2 Tencent AI Lab34±2local~9.4GB2024-01

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

$ what it's good at

text to video▆▄▃▄▆▅█▃▂▂▂▄ #54 of 68 · 98% of its score · 3 tests
testresultvs the others in its rundifficultycountswhy
GenAI-Arena text-to-video (anonymous Elo)1001.9 ✎▃▃···▅▃·▃▅·█ beats 4/10avg 1229▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 49%sharp
AIGVE-60K human perception MOS (LOVE)56.1% ✎▄▄·█··█▄▄▄▄▄ beats 8/10easy · avg 73%▇▇▇▇▇▇▇▇▇▇▇▇ 31%at its level
AIGVE-60K human text-video correspondence MOS (LOVE)50.7% ✎▄▄·█▄█▄··▄▄▄ beats 6/10easy · avg 70%▇▇▇▇▇▇▇ 18%at its level

$ where to run it

Other ways to run it

  1. Run it as published · needs ~9.3 GB, NVIDIA only · from huggingface.co/rhymes-ai (official)
    python -c "import torch;from diffusers import AutoencoderKLAllegro,AllegroPipeline;from diffusers.utils import export_to_video;v=AutoencoderKLAllegro.from_pretrained('rhymes-ai/Allegro',subfolder='vae',torch_dtype=torch.float32);p=AllegroPipeline.from_pretrained('rhymes-ai/Allegro',vae=v,torch_dtype=torch.bfloat16);p.enable_sequential_cpu_offload();p.vae.enable_tiling();export_to_video(p('A seaside harbor with bright sunlight and sparkling seawater, with many boats in the water',guidance_scale=7.5,max_sequence_length=512,num_inference_steps=100).frames[0],'allegro.mp4',fps=15)"

Official weights and documentation: huggingface.co/rhymes-ai ↗

$ how the score was worked out · GenAI-Arena (TIGER-Lab) text-… 49%, LOVE: Benchmarking and Evalua… 48%

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
GenAI-Arena (TIGER-Lab) text-to-video leaderboard1▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 49%MIT
LOVE: Benchmarking and Evaluating Text-to-Video Generation and Video-to-Text Interpretation (AIGVE-60K)2▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 48%CC-BY-4.0
prior · pull toward the middle▇ 2%
weighted average= 35.0

$ 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