← All comparisons
Gemini 3.1 Pro Preview vs Claude 4.5 Sonnet (Reasoning)
Google vs Anthropic — side-by-side benchmark comparison
| Gemini 3.1 Pro Preview | Claude 4.5 Sonnet (Reasoning) | |
|---|---|---|
| Intelligence Index | 57.2 | 43.0 |
| Coding Index | 55.5 | 38.6 |
| Math Index | — | 88.0 |
| Output speed (tok/s) | 136.8 | 55.0 |
| Blended price ($/1M) | $4.50 | $6.56 |
| Time to first token (s) | 25.68s | 7.02s |
| aime | — | — |
| aime 25 | — | 88.0% |
| artificial analysis coding index | 55.50 | 38.60 |
| artificial analysis intelligence index | 57.20 | 43.00 |
| artificial analysis math index | — | 88.00 |
| gpqa | 94.1% | 83.4% |
| hle | 44.7% | 17.3% |
| ifbench | 77.1% | 57.3% |
| lcr | 72.7% | 65.7% |
| livecodebench | — | 71.4% |
| math 500 | — | — |
| mmlu pro | — | 87.5% |
| scicode | 58.9% | 44.7% |
| tau2 | 95.6% | 78.1% |
| terminalbench hard | 53.8% | 35.6% |
Benchmark data from Artificial Analysis.