AssemblyAI: universal
1 benchmarks
Leaderboard position
#532
Overall rank (of 1121 models)
1031
Elo rating
1
Benchmarks
Top 47.5%
Percentile
In AssemblyAI (of 2 models)
Above: inclusionAI: ui-venus-ground-72b #531
Below: luma-labs: luma-uni-1 #533
Benchmark highlights
Best benchmark
-(Artificial Analysis Word Error Rate Index: AgentTalk)
-3.3
Worst benchmark
-(Artificial Analysis Word Error Rate Index: AgentTalk)
-3.3
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 | ~89.45 | 50% | — |
| AIME 2025 | ~71.9 | 50% | — |
| AIME 2026 | ~76.91 | 50% | — |
| HMMT Feb 2026 | ~68.43 | 50% | — |
| GPQA Diamond | ~61.97 | 50% | — |
| MMLU-Pro | ~65.55 | 50% | — |
| Humanity's Last Exam (no tools) | ~14.78 | 50% | — |
| IFEval | ~84.69 | 50% | — |
| IFBench | ~56.26 | 50% | — |
| HumanEval | ~78.84 | 50% | — |
| MBPP | ~83.75 | 50% | — |
| Terminal-Bench v2.1 | ~27.63 | 50% | — |
| MMLU | ~72.22 | 50% | — |
| SWE-bench Verified | ~44.73 | 50% | — |
| SWE-bench Pro | ~23.3 | 50% | — |
| SciCode | ~29.68 | 50% | — |
| AA-LCR | ~46.55 | 50% | — |
| Artificial Analysis Intelligence Index v5.0.1 | ~12 | 50% | — |
| Humanity's Last Exam (with tools) | ~20.05 | 50% | — |
| GSM8K | ~89.46 | 50% | — |
| -(CoderPPL) | ~-1.69 | 50% | — |
| LiveCodeBench v6 | ~59.96 | 50% | — |
| MMMU-Pro | ~56.27 | 50% | — |
| MATH | ~68.17 | 50% | — |
| BBH | ~69.25 | 50% | — |