← All comparisons
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) vs Mi:dm K 2.5 Pro Preview
Anthropic vs Korea Telecom — side-by-side benchmark comparison
| Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) | Mi:dm K 2.5 Pro Preview | |
|---|---|---|
| Intelligence Index | 51.7 | — |
| Coding Index | 50.9 | 11.9 |
| Math Index | — | 78.7 |
| Output speed (tok/s) | 68.2 | 0.0 |
| Blended price ($/1M) | $6.56 | $0.00 |
| Time to first token (s) | 55.35s | 0.00s |
| aime | — | — |
| aime 25 | — | 78.7% |
| artificial analysis coding index | 50.90 | 11.90 |
| artificial analysis intelligence index | 51.70 | — |
| artificial analysis math index | — | 78.70 |
| gpqa | 87.5% | 72.2% |
| hle | 30.0% | 8.8% |
| ifbench | 56.6% | 45.6% |
| lcr | 70.7% | 11.0% |
| livecodebench | — | 57.6% |
| math 500 | — | — |
| mmlu pro | — | 81.3% |
| scicode | 46.8% | 29.7% |
| tau2 | 75.7% | 49.4% |
| terminalbench hard | 53.0% | 3.0% |
Benchmark data from Artificial Analysis.