multimodal benchmark
MusicCaps
MusicCaps is a dataset composed of 5,521 music examples, each labeled with an English aspect list and a free text caption written by musicians. The dataset contains 10-second music clips from AudioSet paired with rich textual descriptions that capture sonic qualities and musical elements like genre, mood, tempo, instrumentation, and rhythm. Created to support research in music-text understanding and generation tasks.
Updated Aug 7, 2026
Published models1
Registry coverage1
MetricScore
EvidenceB
MusicCaps leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AC | 32.8% | 100.0% | 1 | C |
MusicCaps highlights
The top published results on this benchmark's own scale.
What is MusicCaps?
Definition and scoring fields from the benchmark registry.
MusicCaps is a dataset composed of 5,521 music examples, each labeled with an English aspect list and a free text caption written by musicians. The dataset contains 10-second music clips from AudioSet paired with rich textual descriptions that capture sonic qualities and musical elements like genre, mood, tempo, instrumentation, and rhythm. Created to support research in music-text understanding and generation tasks.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MusicCaps
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- musiccaps|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MusicCaps.
Which model scores highest on MusicCaps?
Qwen2.5-Omni-7B is currently ranked first with 32.8%.
What does MusicCaps measure?
MusicCaps is a dataset composed of 5,521 music examples, each labeled with an English aspect list and a free text caption written by musicians. The dataset contains 10-second music clips from AudioSet paired with rich textual descriptions that capture sonic qualities and musical elements like genre, mood, tempo, instrumentation, and rhythm. Created to support research in music-text understanding and generation tasks.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.