Largest Open-Source Human Video Preference Dataset Released: 300K Annotations, 15 SOTA Models Ranked
_akhaliq · x · 2026-09-29
The datapointai team released what they call the largest open-source human video preference dataset and doubled its data grant to $2M.
- Scale: 300K+ preference annotations by real people
- Coverage: 15 SOTA video generation models ranked, including Seedance 2.5, Omni 1.1, Wan 3, and Flux Video 3
- Dimensions: 8 categories covering ads, camera movement, physics plausibility, and more
The dataset and benchmark are open-sourced with a frontier plot, enabling human-preference-based evaluation and training of video generation models.
More from Multimodal
- User Shares Coral Reef Creation Labeled 'Qwen 4 Max (xhigh)', an Unannounced Model Name — intellectronica · 2026-11-16
- Hand-editing 0.2% of weights plus 100 CPU-minutes turns an English TTS model into an 18MB Sanskrit one — kastnerkyle · 2026-11-16
- Zyphra open-sources textnorm tool, lifting TTS structured-data accuracy by up to 22 points — kastnerkyle · 2026-11-16
- Testing Qwen Image 2.1 and Minimax H3 on anime-to-photo conversion and film photo restoration — Ant_6431 · 2026-11-16
- Second take of a 'Cubee run' video generated with Minimax H3 — JahJedi · 2026-11-16
- Non-artist builds 14-chapter hand-painted storybook game with Fable 5.2 in 4 hours — jradoff · 2026-11-16