IBM

Granite 4.1 8B

Apr 29, 2026Open weightsHugging Face

Compare

Intelligence

6.6

#598 of 744Rank 598 of 744 measured models

Coding

9.5

#275 of 298Rank 275 of 298 measured models

Output speed

126 t/s

#123 of 403Rank 123 of 403 measured models

Price/1M

$0.063

#61 of 659Rank 61 of 659 measured models

Context window

131K

#738 of 1199Rank 738 of 1199 measured models

Granite 4.1 8B scores 6.6 on the Artificial Analysis Intelligence Index, ranking 598th of 744 measured models. The closest models on this index are DeepSeek-V2.5 (6.6), Gemini 2.0 Flash Thinking Experimental (Dec '24) (6.6), and Llama 3.1 Instruct 70B (6.6). At $0.063 per million tokens (3:1 input/output blended price), it is cheaper than Llama 3.1 Instruct 70B ($0.56). With an output speed of 126 t/s, it is faster than Llama 3.1 Instruct 70B (36.7 t/s).