BASS - Section Structural Segmentation: leaderboard

Metric: Label-matched intersection over union (%) on the 224 section segmentation questions: return the timestamps of one section type or one named section (free-form answers); BASS music understanding benchmark, audio language models get a text prompt (one of ten paraphrases) and a full commercial music recording from BASS's songs; raw score before chance normalization; higher is better. Source: arxiv.org. Saturation forecast: Around April 2027. 14 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro22.98#145
2Voxtral-Small-24B-25078.84#565
3Phi-4 Multimodal Instruct0.16#896
4SALMONN-7B0.12#1274

No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.

Interactive version: theaggregate.ai/benchmark?slug=bass-section-structural-segmentation · How It Works · Data refreshed daily, snapshot 2026-10-11.