boogiebench: leaderboard
Elo-ranked LLM music-composition arena where text models write Strudel JavaScript compositions from music prompts and users vote on the generated tracks.
Metric: Elo Rating. Source: www.boogiebench.com. Saturation forecast: Around April 2027. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 1658 | #63 |
| 2 | Gemini 3 Pro (Preview) | 1658 | #66 |
| 3 | GPT-5 | 1634 | #98 |
| 4 | GPT-5.2 | 1624 | #110 |
| 5 | GPT-5.4 | 1617 | #79 |
| 6 | Claude Opus 4.5 | 1606 | #83 |
| 7 | Claude Sonnet 4.5 | 1591 | #142 |
| 8 | Kimi K2 0905 | 1435 | #288 |
| 9 | GLM-4.6 | 1426 | #252 |
| 10 | DeepSeek V3.1 Terminus | 1420 | #220 |
| 11 | Gemini 2.5 Pro | 1401 | #152 |
| 12 | GPT-OSS-120B | 1397 | #341 |
| 13 | MiniMax-M2 | 1314 | #317 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=boogiebench · How It Works · Data refreshed daily, snapshot 2026-10-20.