CMI-RewardBench - Music Arena Musicality: leaderboard
Metric: Pairwise accuracy (%) against the human preference on 1,340 Music Arena comparison pairs (interaction logs to December 2025, failed generations and tie or both-bad votes removed), treated as musicality labels; the model is given the two clips (with the instruction) and must select the preferred one; the zero-shot audio LLMs are prompted to choose, the reward models compare their scalar scores; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 69.75 | #145 |
| 2 | Gemini 3 Pro | 68.85 | #77 |
| 3 | Gemini 2.5 Flash | 64.12 | #237 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=cmi-rewardbench-music-arena-musicality · How It Works · Data refreshed daily, snapshot 2026-10-11.