Google: veo-3-audio
1 benchmarks
Leaderboard position
#553
Overall rank (of 1121 models)
990
Elo rating
1
Benchmarks
Top 49.3%
Percentile
In Google (of 72 models)
Above: wan-ai: wan2.5-preview #552
Below: Qwen: qwen3-vl-235b-a22b-instruct #554
Benchmark highlights
Best benchmark
Text-to-Video Arena (Arena.ai) (Sep 21, 2026)
1340
Worst benchmark
Text-to-Video Arena (Arena.ai) (Sep 21, 2026)
1340
Nearest competitor from another company
The model from another company with the most similar benchmark scores.
appeared as nearest in 1 of 1 benchmarks
1/1
100.0%
Benchmarks
Predicted scores
Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.
| Benchmark | Predicted | Confidence | Based on |
|---|---|---|---|
| MATH-500 | ~88.74 | 50% | — |
| AIME 2025 | ~70.55 | 50% | — |
| AIME 2026 | ~76.12 | 50% | — |
| HMMT Feb 2026 | ~67.64 | 50% | — |
| GPQA Diamond | ~61.07 | 50% | — |
| MMLU-Pro | ~64.55 | 50% | — |
| Humanity's Last Exam (no tools) | ~14.05 | 50% | — |
| IFEval | ~84.13 | 50% | — |
| IFBench | ~55.58 | 50% | — |
| HumanEval | ~77.65 | 50% | — |
| MBPP | ~82.19 | 50% | — |
| Terminal-Bench v2.1 | ~26.2 | 50% | — |
| MMLU | ~71.17 | 50% | — |
| SWE-bench Verified | ~43.45 | 50% | — |
| SWE-bench Pro | ~22.1 | 50% | — |
| SciCode | ~29.04 | 50% | — |
| AA-LCR | ~45.75 | 50% | — |
| Artificial Analysis Intelligence Index v5.0.1 | ~12 | 50% | — |
| Humanity's Last Exam (with tools) | ~19.08 | 50% | — |
| GSM8K | ~87.92 | 50% | — |
| -(CoderPPL) | ~-1.69 | 50% | — |
| LiveCodeBench v6 | ~59.03 | 50% | — |
| MMMU-Pro | ~55.5 | 50% | — |
| MATH | ~66.45 | 50% | — |
| BBH | ~68.34 | 50% | — |