← All comparisons
Gemma 3 12B Instruct vs Claude 4.1 Opus (Reasoning)
Google vs Anthropic — side-by-side benchmark comparison
| Gemma 3 12B Instruct | Claude 4.1 Opus (Reasoning) | |
|---|---|---|
| Intelligence Index | 8.8 | 42.0 |
| Coding Index | 6.3 | 36.5 |
| Math Index | 18.3 | 80.3 |
| Output speed (tok/s) | 0.0 | 44.5 |
| Blended price ($/1M) | $0.14 | $32.81 |
| Time to first token (s) | 0.00s | 8.55s |
| aime | 22.0% | — |
| aime 25 | 18.3% | 80.3% |
| artificial analysis coding index | 6.30 | 36.50 |
| artificial analysis intelligence index | 8.80 | 42.00 |
| artificial analysis math index | 18.30 | 80.30 |
| gpqa | 34.9% | 80.9% |
| hle | 4.8% | 11.9% |
| ifbench | 36.7% | 55.4% |
| lcr | 6.7% | 66.3% |
| livecodebench | 13.7% | 65.4% |
| math 500 | 85.3% | — |
| mmlu pro | 59.5% | 88.0% |
| scicode | 17.4% | 40.9% |
| tau2 | 10.8% | 71.4% |
| terminalbench hard | 0.8% | 34.3% |
Benchmark data from Artificial Analysis.