Across 10 shared benchmarks, Claude Sonnet 4 scores higher on 7 and Granite 4.2 30B on 3. The widest gap is τ³-Bench, where Granite 4.2 30B scores 68 against 13.8.
| Benchmark | Claude Sonnet 4 | Granite 4.2 30B |
|---|---|---|
| gdpval | 31.2 | 1225 |
| GPQA Diamond | 78 | 66.4 |
| LiveCodeBench v6 | 48.5 | 75.8 |
| MMLU-Pro | 84 | 77.6 |
| scicode | 40 | 38.8 |
| SWE-bench Multilingual | 56.9 | 41.9 |
| SWE-bench Pro | 42.7 | 33.3 |
| SWE-bench Verified | 80.2 | 57 |
| Terminal-Bench 2.1 | 36.3 | 29.2 |
| τ³-Bench | 13.8 | 68 |
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.