Which is better, Claude Opus 4.6 or Granite 4.2 30B?

Refresh Models —Deals —Regulatory —Sources —

Questions this page answers

Which is better, Claude Opus 4.6 or Granite 4.2 30B?: Across 10 shared benchmarks, Claude Opus 4.6 scores higher on 9 and Granite 4.2 30B on 1. The widest gap is SWE-bench Multilingual, where Claude Opus 4.6 scores 77.8 against 41.9.

Claude Opus 4.6 vs Granite 4.2 30B

Across 10 shared benchmarks, Claude Opus 4.6 scores higher on 9 and Granite 4.2 30B on 1. The widest gap is SWE-bench Multilingual, where Claude Opus 4.6 scores 77.8 against 41.9.

AnthropicvsIBM10 shared benchmarks9–1 head-to-head

Benchmark	Claude Opus 4.6	Granite 4.2 30B
AIME25 no tools	95.6	89.2
gdpval	55.9	1225
GPQA Diamond	91.3	66.4
LiveCodeBench v6	88.8	75.8
MMLU-Pro	89.1	77.6
scicode	52	38.8
SWE-bench Multilingual	77.8	41.9
SWE-bench Pro	57.3	33.3
SWE-bench Verified	80.8	57
τ³-Bench	72.4	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.