← All comparisons
K2-V2 (high) vs Claude 4.1 Opus (Non-reasoning)
MBZUAI Institute of Foundation Models vs Anthropic — side-by-side benchmark comparison
| K2-V2 (high) | Claude 4.1 Opus (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 20.6 | 36.0 |
| Coding Index | 16.1 | — |
| Math Index | 78.3 | — |
| Output speed (tok/s) | 0.0 | 44.7 |
| Blended price ($/1M) | $0.00 | $32.81 |
| Time to first token (s) | 0.00s | 1.63s |
| aime | — | — |
| aime 25 | 78.3% | — |
| artificial analysis coding index | 16.10 | — |
| artificial analysis intelligence index | 20.60 | 36.00 |
| artificial analysis math index | 78.30 | — |
| gpqa | 68.1% | — |
| hle | 9.8% | — |
| ifbench | 60.1% | — |
| lcr | 33.3% | — |
| livecodebench | 69.4% | — |
| math 500 | — | — |
| mmlu pro | 78.6% | — |
| scicode | 28.6% | — |
| tau2 | 27.8% | — |
| terminalbench hard | 9.8% | — |
Benchmark data from Artificial Analysis.