Which is better, GPT-5 or Granite 4.2 30B?

Refresh Models —Deals —Regulatory —Sources —

Questions this page answers

Which is better, GPT-5 or Granite 4.2 30B?: Across 11 shared benchmarks, GPT-5 scores higher on 9 and Granite 4.2 30B on 2. The widest gap is τ³-Bench, where Granite 4.2 30B scores 68 against 19.6.

GPT-5 vs Granite 4.2 30B

Across 11 shared benchmarks, GPT-5 scores higher on 9 and Granite 4.2 30B on 2. The widest gap is τ³-Bench, where Granite 4.2 30B scores 68 against 19.6.

OpenAIvsIBM11 shared benchmarks9–2 head-to-head

Benchmark	GPT-5	Granite 4.2 30B
AIME25 no tools	94.6	89.2
gdpval	32.5	1225
GPQA Diamond	87.3	66.4
LiveCodeBench v6	87	75.8
MMLU-Pro	87.5	77.6
scicode	43	38.8
SWE-bench Multilingual	55.3	41.9
SWE-bench Pro	41.8	33.3
SWE-bench Verified	74.9	57
Terminal-Bench 2.1	35.2	29.2
τ³-Bench	19.6	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.