← All comparisons
Cogito v2.1 (Reasoning) vs Claude 4.5 Sonnet (Reasoning)
Deep Cogito vs Anthropic — side-by-side benchmark comparison
| Cogito v2.1 (Reasoning) | Claude 4.5 Sonnet (Reasoning) | |
|---|---|---|
| Intelligence Index | — | 43.0 |
| Coding Index | 24.8 | 38.6 |
| Math Index | 72.7 | 88.0 |
| Output speed (tok/s) | 80.7 | 55.0 |
| Blended price ($/1M) | $1.25 | $6.56 |
| Time to first token (s) | 0.51s | 7.02s |
| aime | — | — |
| aime 25 | 72.7% | 88.0% |
| artificial analysis coding index | 24.80 | 38.60 |
| artificial analysis intelligence index | — | 43.00 |
| artificial analysis math index | 72.70 | 88.00 |
| gpqa | 76.8% | 83.4% |
| hle | 11.0% | 17.3% |
| ifbench | 46.3% | 57.3% |
| lcr | 21.7% | 65.7% |
| livecodebench | 68.8% | 71.4% |
| math 500 | — | — |
| mmlu pro | 84.9% | 87.5% |
| scicode | 41.0% | 44.7% |
| tau2 | — | 78.1% |
| terminalbench hard | 16.7% | 35.6% |
Benchmark data from Artificial Analysis.