UniVerseBench
Models answer questions about music across more than 38 cultural and linguistic groups. The authors report accuracy by language and skill.
- Made by
- Ziya Zhou and colleagues
- Released
- August 2026
- Kind
- Audio understanding and voice
- Level
- University
- Scored
- Checked automatically
- Answers
- Multiple choice
- Tasks
- All public
- Measures
- Capability
- Subject
- Arts
- Language
- Many languages
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Audio and speech, Many languages
- Strongest model tested
-
Gemini 3 Flash #60 on the capability index
- Last model added
- March 2026
- Scores refreshed
- November 20th, 2026
Qwen3.5-Omni-Plus leads with 74.4% among models released by March 2026 Qwen3.5-Omni-Plus leads with 74.4% among models released by March 2026
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. This view shows only models we track, each at its best.
The newest model with a score here came out in March 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1
Qwen3.5-Omni-Plus 74.4%
- 2
Gemini 3 Flash 70.8%
- 3 Music Flamingo 56%
- 4 Qwen3-Omni-30B-A3B (SFT, w/ think) 53.4%
- 5 Qwen2.5-Omni-7B-Instruct (LR Phase-b) 48.8%
- 6
Qwen3-Omni-30B-A3B 47.5%
- 7 Kimi-Audio-7B-Instruct 46.2%
- 8 MiDashengLM 42.5%
- 9
Qwen2.5-Omni-7B 33.9%
Source: Official UniVerseBench leaderboard
- 1
Qwen3.5-Omni-Plus 74.4%
- 2
Gemini 3 Flash 70.8%
- 3
Qwen3-Omni-30B-A3B 47.5%
- 4 Kimi-Audio-7B-Instruct 46.2%
- 5
Qwen2.5-Omni-7B 33.9%
Source: Official UniVerseBench leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 74.4% | Mar 2026 | Official leaderboard | |
| 2 | | 70.8% | Dec 2025 | Official leaderboard | |
| 3 | Music Flamingo | 56% | Official leaderboard | ||
| 4 | Qwen3-Omni-30B-A3B (SFT, w/ think) | 53.4% | Official leaderboard | ||
| 5 | Qwen2.5-Omni-7B-Instruct (LR Phase-b) | 48.8% | Official leaderboard | ||
| 6 | | 47.5% | Sep 2025 | Official leaderboard | |
| 7 | Kimi-Audio-7B-Instruct | 46.2% | Apr 2025 | Official leaderboard | |
| 8 | MiDashengLM | 42.5% | Official leaderboard | ||
| 9 | | 33.9% | Mar 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 74.4% | Mar 2026 | Official leaderboard | |
| 2 | | 70.8% | Dec 2025 | Official leaderboard | |
| 3 | | 47.5% | Sep 2025 | Official leaderboard | |
| 4 | Kimi-Audio-7B-Instruct | 46.2% | Apr 2025 | Official leaderboard | |
| 5 | | 33.9% | Mar 2025 | Official leaderboard |
Rankings by part
We read 8 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Chinese, Korean and Other languages. Qwen3.5-Omni-Plus leads 6 of them, and Gemini 3 Flash leads the other 2. Arrows show how far a model moved from its overall rank among the models we track.
Chinese
5 models we track have a score here.
- 1
Qwen3.5-Omni-Plus 71%
- 2
Gemini 3 Flash 67.6%
- 3
Qwen3-Omni-30B-A3B 44.3%
- 4 Kimi-Audio-7B-Instruct 37.6%
- 5
Qwen2.5-Omni-7B 29.2%
Korean
5 models we track have a score here.
- 1
Gemini 3 Flash ↑1 77.6%
- 2
Qwen3.5-Omni-Plus ↓1 75.5%
- 3 Kimi-Audio-7B-Instruct ↑1 56.4%
- 4
Qwen3-Omni-30B-A3B ↓1 52.6%
- 5
Qwen2.5-Omni-7B 45.2%
Other languages
5 models we track have a score here.
- 1
Qwen3.5-Omni-Plus 75.7%
- 2
Gemini 3 Flash 69.5%
- 3
Qwen3-Omni-30B-A3B 47.3%
- 4 Kimi-Audio-7B-Instruct 46.8%
- 5
Qwen2.5-Omni-7B 32%
Feature Extraction
5 models we track have a score here.
- 1
Qwen3.5-Omni-Plus 73.6%
- 2
Gemini 3 Flash 68.7%
- 3
Qwen3-Omni-30B-A3B 44.9%
- 4 Kimi-Audio-7B-Instruct 44.7%
- 5
Qwen2.5-Omni-7B 32.5%
Causal Reasoning
5 models we track have a score here.
- 1
Qwen3.5-Omni-Plus 80.1%
- 2
Gemini 3 Flash 73.3%
- 3
Qwen3-Omni-30B-A3B 58.3%
- 4 Kimi-Audio-7B-Instruct 47.6%
- 5
Qwen2.5-Omni-7B 38.3%
Music Pattern Recognition
5 models we track have a score here.
- 1
Qwen3.5-Omni-Plus 72.6%
- 2
Gemini 3 Flash 71.3%
- 3
Qwen3-Omni-30B-A3B 51.4%
- 4 Kimi-Audio-7B-Instruct 48.6%
- 5
Qwen2.5-Omni-7B 35.7%
Comparison Analysis
5 models we track have a score here.
- 1
Gemini 3 Flash ↑1 84.9%
- 2
Qwen3.5-Omni-Plus ↓1 69%
- 3 Kimi-Audio-7B-Instruct ↑1 57.3%
- 4
Qwen3-Omni-30B-A3B ↓1 37.5%
- 5
Qwen2.5-Omni-7B 32.8%
Error Detection
5 models we track have a score here.
- 1
Qwen3.5-Omni-Plus 77.1%
- 2
Gemini 3 Flash 55.6%
- 3 Kimi-Audio-7B-Instruct ↑1 51.4%
- 4
Qwen3-Omni-30B-A3B ↓1 51.4%
- 5
Qwen2.5-Omni-7B 37.1%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official UniVerseBench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official UniVerseBench 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
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 8 months ago, in March 2026, for Qwen3.5-Omni-Plus. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the official leaderboard (9 entries).
- Pass Models scored 5 models we track, plus 4 other entries on its leaderboard.
- Warning Newest score The newest came 8 months ago, in March 2026, for Qwen3.5-Omni-Plus.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 3 Flash (#60).
- Pass Headroom The best is 74.4%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on 143.89.224.8: 5 rows for models we track and 4 other entries at the last read.
- 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.