← All comparisons
Gemini 2.5 Pro vs Grok 3 mini Reasoning (high)
Google vs xAI — side-by-side benchmark comparison
| Gemini 2.5 Pro | Grok 3 mini Reasoning (high) | |
|---|---|---|
| Intelligence Index | 34.6 | 32.1 |
| Coding Index | 32.0 | 25.2 |
| Math Index | 87.7 | 84.7 |
| Output speed (tok/s) | 140.4 | 56.8 |
| Blended price ($/1M) | $3.44 | $0.35 |
| Time to first token (s) | 20.87s | 0.42s |
| aime | 88.7% | 93.3% |
| aime 25 | 87.7% | 84.7% |
| artificial analysis coding index | 32.00 | 25.20 |
| artificial analysis intelligence index | 34.60 | 32.10 |
| artificial analysis math index | 87.70 | 84.70 |
| gpqa | 84.4% | 79.1% |
| hle | 21.1% | 11.1% |
| ifbench | 48.7% | 45.9% |
| lcr | 66.0% | 50.3% |
| livecodebench | 80.1% | 69.6% |
| math 500 | 96.7% | 99.2% |
| mmlu pro | 86.2% | 82.8% |
| scicode | 42.8% | 40.6% |
| tau2 | 54.1% | 90.4% |
| terminalbench hard | 26.5% | 17.4% |
Benchmark data from Artificial Analysis.