IBM

Granite 4.2 30B

Aug 25, 2026ReasoningOpen weightsHugging Face

Compare

Intelligence

12.7

#356 of 746Rank 356 of 746 measured models

Coding

29.9

#185 of 298Rank 185 of 298 measured models

Output speed

75.5 t/s

#245 of 403Rank 245 of 403 measured models

Price/1M

$0.282

#205 of 659Rank 205 of 659 measured models

Context window

131K

#740 of 1202Rank 740 of 1202 measured models

Cost per task

$0.0717

#43 of 193Rank 43 of 193 measured models

Granite 4.2 30B scores 12.7 on the Artificial Analysis Intelligence Index, ranking 356th of 746 measured models. The closest models on this index are GPT-4.1 (12.7), Kimi K2 (12.7), and Qwen3 235B A22B 2507 (Reasoning) (12.7). At $0.282 per million tokens (3:1 input/output blended price), it is cheaper than GPT-4.1 ($3.50), Kimi K2 ($1.00), and Qwen3 235B A22B 2507 (Reasoning) ($0.747). With an output speed of 75.5 t/s, it is faster than Kimi K2 (54.1 t/s) and Qwen3 235B A22B 2507 (Reasoning) (43.9 t/s), slower than GPT-4.1 (155 t/s).