← All comparisons
Cogito v2.1 (Reasoning) vs Qwen3 Max (Preview)
Deep Cogito vs Alibaba — side-by-side benchmark comparison
| Cogito v2.1 (Reasoning) | Qwen3 Max (Preview) | |
|---|---|---|
| Intelligence Index | — | 26.1 |
| Coding Index | 24.8 | 25.5 |
| Math Index | 72.7 | 75.0 |
| Output speed (tok/s) | 80.7 | 50.8 |
| Blended price ($/1M) | $1.25 | $2.40 |
| Time to first token (s) | 0.51s | 1.99s |
| aime | — | — |
| aime 25 | 72.7% | 75.0% |
| artificial analysis coding index | 24.80 | 25.50 |
| artificial analysis intelligence index | — | 26.10 |
| artificial analysis math index | 72.70 | 75.00 |
| gpqa | 76.8% | 76.4% |
| hle | 11.0% | 9.3% |
| ifbench | 46.3% | 48.0% |
| lcr | 21.7% | 39.7% |
| livecodebench | 68.8% | 65.1% |
| math 500 | — | — |
| mmlu pro | 84.9% | 83.8% |
| scicode | 41.0% | 37.0% |
| tau2 | — | 32.7% |
| terminalbench hard | 16.7% | 19.7% |
Benchmark data from Artificial Analysis.