← All comparisons
gpt-oss-120b (low) vs Granite 4.0 H Small
OpenAI vs IBM — side-by-side benchmark comparison
| gpt-oss-120b (low) | Granite 4.0 H Small | |
|---|---|---|
| Intelligence Index | 24.5 | 10.8 |
| Coding Index | 15.5 | 8.5 |
| Math Index | 66.7 | 13.7 |
| Output speed (tok/s) | 370.0 | 401.2 |
| Blended price ($/1M) | $0.26 | $0.11 |
| Time to first token (s) | 0.49s | 8.75s |
| aime | — | — |
| aime 25 | 66.7% | 13.7% |
| artificial analysis coding index | 15.50 | 8.50 |
| artificial analysis intelligence index | 24.50 | 10.80 |
| artificial analysis math index | 66.70 | 13.70 |
| gpqa | 67.2% | 41.6% |
| hle | 5.2% | 3.7% |
| ifbench | 58.3% | 31.5% |
| lcr | 43.7% | 9.0% |
| livecodebench | 70.7% | 25.1% |
| math 500 | — | — |
| mmlu pro | 77.5% | 62.4% |
| scicode | 36.0% | 20.9% |
| tau2 | 45.0% | 17.3% |
| terminalbench hard | 5.3% | 2.3% |
Benchmark data from Artificial Analysis.