Google: gemini-decision-1
3 benchmarks
Leaderboard position
#187
Overall rank (of 1121 models)
1932
Elo rating
3
Benchmarks
Top 16.7%
Percentile
In Google (of 72 models)
Above: recraft: recraft-v4.1-utility #186
Below: inworld: realtime-tts-2-research-preview #188
Benchmark highlights
Best benchmark
When2Call
79.9
Worst benchmark
Video-MME (w/o subs)
69.1
Nearest competitor from another company
The model from another company with the most similar benchmark scores.
appeared as nearest in 1 of 3 benchmarks
| Model | Votes | % of benchmarks |
|---|---|---|
| typesafe: jev-1.14 | 1 | 33.3% |
| biomistral: biomistral-7b | 1 | 33.3% |
| OpenBMB: minicpm-v-4.6 | 1 | 33.3% |
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 | ~99.2 | 50% | — |
| AIME 2025 | ~100 | 50% | — |
| AIME 2026 | ~94.34 | 50% | — |
| HMMT Feb 2026 | ~85.92 | 50% | — |
| GPQA Diamond | ~81.79 | 50% | — |
| MMLU-Pro | ~87.66 | 50% | — |
| Humanity's Last Exam (no tools) | ~30.81 | 50% | — |
| IFEval | ~96.75 | 50% | — |
| IFBench | ~71.17 | 50% | — |
| HumanEval | ~95.9 | 50% | — |
| MBPP | ~93.7 | 50% | — |
| Terminal-Bench v2.1 | ~58.85 | 50% | — |
| MMLU | ~92.5 | 50% | — |
| SWE-bench Verified | ~73 | 50% | — |
| SWE-bench Pro | ~49.66 | 50% | — |
| SciCode | ~43.75 | 50% | — |
| AA-LCR | ~64.15 | 50% | — |
| Artificial Analysis Intelligence Index v5.0.1 | ~12.96 | 50% | — |
| Humanity's Last Exam (with tools) | ~41.46 | 50% | — |
| GSM8K | ~99.6 | 50% | — |
| -(CoderPPL) | ~-1.69 | 50% | — |
| LiveCodeBench v6 | ~80.55 | 50% | — |
| MMMU-Pro | ~73.18 | 50% | — |
| MATH | ~97.9 | 50% | — |
| BBH | ~88.4 | 50% | — |