← All comparisons
Cogito v2.1 (Reasoning) vs GPT-5 (low)
Deep Cogito vs OpenAI — side-by-side benchmark comparison
| Cogito v2.1 (Reasoning) | GPT-5 (low) | |
|---|---|---|
| Intelligence Index | — | 39.2 |
| Coding Index | 24.8 | 30.7 |
| Math Index | 72.7 | 83.0 |
| Output speed (tok/s) | 80.7 | 88.9 |
| Blended price ($/1M) | $1.25 | $3.44 |
| Time to first token (s) | 0.51s | 8.32s |
| aime | — | 83.0% |
| aime 25 | 72.7% | 83.0% |
| artificial analysis coding index | 24.80 | 30.70 |
| artificial analysis intelligence index | — | 39.20 |
| artificial analysis math index | 72.70 | 83.00 |
| gpqa | 76.8% | 80.8% |
| hle | 11.0% | 18.4% |
| ifbench | 46.3% | 66.6% |
| lcr | 21.7% | 58.7% |
| livecodebench | 68.8% | 76.3% |
| math 500 | — | 98.7% |
| mmlu pro | 84.9% | 86.0% |
| scicode | 41.0% | 39.1% |
| tau2 | — | 84.2% |
| terminalbench hard | 16.7% | 26.5% |
Benchmark data from Artificial Analysis.