TMD-Bench - Video Instruction Following: leaderboard
Metric: Video instruction following (0-100): the Gemini 3 Pro judgment of how well the generated dance video matches the content the prompt implies, normalized to 0-1 by the paper and reported here times 100, on the 100 prompts of the TMD-Bench test set (text-driven joint generation of music and a dance video); higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Sora 2 | 91 |
| 2 | Veo 3 | 86 |
| 3 | Wan 2.6 | 76 |
| 4 | Kling 2.6 | 74 |
| 5 | Seedance 1.5 pro | 71 |
| 6 | Ovi | 58 |
| 7 | JavisDiT | 53 |
| 8 | UniVerse-1 | 52 |
| 9 | LTX-2 | 52 |
Interactive version: theaggregate.ai/benchmark?slug=tmd-bench-video-instruction-following · How It Works · Data refreshed daily, snapshot 2026-10-07.