← All comparisons
K2 Think V2 vs DeepSeek R1 Distill Qwen 1.5B
MBZUAI Institute of Foundation Models vs DeepSeek — side-by-side benchmark comparison
| K2 Think V2 | DeepSeek R1 Distill Qwen 1.5B | |
|---|---|---|
| Intelligence Index | 24.1 | 9.1 |
| Coding Index | 15.5 | — |
| Math Index | — | 22.0 |
| Output speed (tok/s) | 0.0 | 0.0 |
| Blended price ($/1M) | $0.00 | $0.00 |
| Time to first token (s) | 0.00s | 0.00s |
| aime | — | 17.7% |
| aime 25 | — | 22.0% |
| artificial analysis coding index | 15.50 | — |
| artificial analysis intelligence index | 24.10 | 9.10 |
| artificial analysis math index | — | 22.00 |
| gpqa | 71.3% | 9.8% |
| hle | 9.5% | 3.3% |
| ifbench | 62.8% | 13.2% |
| lcr | 52.7% | 0.3% |
| livecodebench | — | 7.0% |
| math 500 | — | 68.7% |
| mmlu pro | — | 26.9% |
| scicode | 33.0% | 6.6% |
| tau2 | 25.4% | — |
| terminalbench hard | 6.8% | — |
Benchmark data from Artificial Analysis.