← All comparisons
Gemma 4 E4B (Reasoning) vs Claude Opus 4.8 (Adaptive Reasoning, Max Effort)
Google vs Anthropic — side-by-side benchmark comparison
| Gemma 4 E4B (Reasoning) | Claude Opus 4.8 (Adaptive Reasoning, Max Effort) | |
|---|---|---|
| Intelligence Index | 18.8 | 61.4 |
| Coding Index | 13.7 | 56.7 |
| Math Index | — | — |
| Output speed (tok/s) | 0.0 | 66.9 |
| Blended price ($/1M) | $0.00 | $10.94 |
| Time to first token (s) | 0.00s | 7.91s |
| aime | — | — |
| aime 25 | — | — |
| artificial analysis coding index | 13.70 | 56.70 |
| artificial analysis intelligence index | 18.80 | 61.40 |
| artificial analysis math index | — | — |
| gpqa | 57.6% | 92.0% |
| hle | 3.7% | 45.7% |
| ifbench | 44.2% | 62.2% |
| lcr | 30.7% | 67.7% |
| livecodebench | — | — |
| math 500 | — | — |
| mmlu pro | — | — |
| scicode | 24.4% | 53.5% |
| tau2 | 20.8% | 94.4% |
| terminalbench hard | 8.3% | 58.3% |
Benchmark data from Artificial Analysis.