BirdSet
Models identify bird species from audio recordings across multiple evaluation datasets.
The benchmark includes recordings with multiple species sounding at once.
- Made by
- Lukas Rauch and colleagues
- Released
- March 2024
- Kind
- Audio understanding and voice
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Prediction
- Tasks
- All public
- Measures
- Capability
- Subject
- Biological sciences
- Interaction
- Single turn
- Items from
- Real work and use, Another benchmark
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Audio and speech, Biology and medicine
DBD-research-group-ConvNeXT-XCL tops the leaderboard with 0.85
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.
- 0.85 DBD-research-group-ConvNeXT-XCL
- 0.85 DBD-research-group-EfficientNet-B1-XCM
- 0.84 DBD-research-group-AST-XCL
- 0.84 DBD-research-group-ConvNeXT-XCM
- 0.84 DBD-research-group-EfficientNet-B1-Dedicated
- 0.84 DBD-research-group-EfficientNet-B1-XCL
- 0.84 Perch
- 0.83 DBD-research-group-AST-Dedicated
- 0.83 DBD-research-group-AST-XCM
- 0.83 DBD-research-group-ConvNeXT-Dedicated
- 0.82 DBD-research-group-EAT-XCL
- 0.80 DBD-research-group-EAT-XCM
- 0.80 DBD-research-group-Wav2Vec2-XCM
- 0.79 DBD-research-group-Wav2Vec2-Dedicated
- 0.78 DBD-research-group-EAT-Dedicated
- 0.78 DBD-research-group-Wav2Vec2-XCL
- 1 DBD-research-group-ConvNeXT-XCL lt, float32 0.85
- 2 DBD-research-group-EfficientNet-B1-XCM XCM, float32 0.85
- 3 DBD-research-group-AST-XCL XCL, float32 0.84
- 4 DBD-research-group-ConvNeXT-XCM XCM, float32 0.84
- 5 DBD-research-group-EfficientNet-B1-Dedicated Dedicated, float32 0.84
- 6 DBD-research-group-EfficientNet-B1-XCL XCL, float32 0.84
- 7 Perch Xeno-Canto (Custom), Other 0.84
- 8 DBD-research-group-AST-Dedicated Dedicated, float32 0.83
- 9 DBD-research-group-AST-XCM XCM, float32 0.83
- 10 DBD-research-group-ConvNeXT-Dedicated Dedicated, float32 0.83
- 11 DBD-research-group-EAT-XCL XCL, Other 0.82
- 12 DBD-research-group-EAT-XCM XCM, Other 0.80
- 13 DBD-research-group-Wav2Vec2-XCM XCM, float32 0.80
- 14 DBD-research-group-Wav2Vec2-Dedicated Dedicated, float32 0.79
- 15 DBD-research-group-EAT-Dedicated Dedicated, Other 0.78
- 16 DBD-research-group-Wav2Vec2-XCL XCL, float32 0.78
Source: Official BirdSet leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | DBD-research-group-ConvNeXT-XCL | 0.85 | lt, float32 | Official leaderboard | |
| 2 | DBD-research-group-EfficientNet-B1-XCM | 0.85 | XCM, float32 | Official leaderboard | |
| 3 | DBD-research-group-AST-XCL | 0.84 | XCL, float32 | Official leaderboard | |
| 4 | DBD-research-group-ConvNeXT-XCM | 0.84 | XCM, float32 | Official leaderboard | |
| 5 | DBD-research-group-EfficientNet-B1-Dedicated | 0.84 | Dedicated, float32 | Official leaderboard | |
| 6 | DBD-research-group-EfficientNet-B1-XCL | 0.84 | XCL, float32 | Official leaderboard | |
| 7 | Perch | 0.84 | Xeno-Canto (Custom), Other | Official leaderboard | |
| 8 | DBD-research-group-AST-Dedicated | 0.83 | Dedicated, float32 | Official leaderboard | |
| 9 | DBD-research-group-AST-XCM | 0.83 | XCM, float32 | Official leaderboard | |
| 10 | DBD-research-group-ConvNeXT-Dedicated | 0.83 | Dedicated, float32 | Official leaderboard | |
| 11 | DBD-research-group-EAT-XCL | 0.82 | XCL, Other | Official leaderboard | |
| 12 | DBD-research-group-EAT-XCM | 0.80 | XCM, Other | Official leaderboard | |
| 13 | DBD-research-group-Wav2Vec2-XCM | 0.80 | XCM, float32 | Official leaderboard | |
| 14 | DBD-research-group-Wav2Vec2-Dedicated | 0.79 | Dedicated, float32 | Official leaderboard | |
| 15 | DBD-research-group-EAT-Dedicated | 0.78 | Dedicated, Other | Official leaderboard | |
| 16 | DBD-research-group-Wav2Vec2-XCL | 0.78 | XCL, float32 | Official leaderboard |
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Audio and video
- Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy.
Qwen 3.8 Max 88.7%
- LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization.
Gemini 4 Argon 91.7%
- MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy.
Qwen 3.8 Max 82.4%
- MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy.
Qwen 3.8 Max 90.8%
- FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better.
Phi-4-multimodal 4.00%
- VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules.
Qwen3.5-Omni-Plus 93.1%
Data audit
All 8 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the official leaderboard (16 entries).
- Pass Models scored 16 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on huggingface.co: 16 entries at the last read, none a model we track.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.