Across 6 shared benchmarks, Claude 3 Opus scores higher on 4 and IBM Granite 4.0 Small on 1, with 1 level. The widest gap is xstest, where Claude 3 Opus scores 92.5 against 80.6.
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.