← All comparisons
Granite 4.0 H Small vs Claude Opus 4.6 (Adaptive Reasoning, Max Effort)
IBM vs Anthropic — side-by-side benchmark comparison
| Granite 4.0 H Small | Claude Opus 4.6 (Adaptive Reasoning, Max Effort) | |
|---|---|---|
| Intelligence Index | 10.8 | 52.9 |
| Coding Index | 8.5 | 48.1 |
| Math Index | 13.7 | — |
| Output speed (tok/s) | 401.2 | 54.8 |
| Blended price ($/1M) | $0.11 | $10.94 |
| Time to first token (s) | 8.75s | 11.69s |
| aime | — | — |
| aime 25 | 13.7% | — |
| artificial analysis coding index | 8.50 | 48.10 |
| artificial analysis intelligence index | 10.80 | 52.90 |
| artificial analysis math index | 13.70 | — |
| gpqa | 41.6% | 89.6% |
| hle | 3.7% | 36.7% |
| ifbench | 31.5% | 53.1% |
| lcr | 9.0% | 70.7% |
| livecodebench | 25.1% | — |
| math 500 | — | — |
| mmlu pro | 62.4% | — |
| scicode | 20.9% | 51.9% |
| tau2 | 17.3% | 92.1% |
| terminalbench hard | 2.3% | 46.2% |
Benchmark data from Artificial Analysis.