← All comparisons

Grok 4.3 (Non-reasoning) vs Granite 4.0 H 1B

xAI vs IBM — side-by-side benchmark comparison

Grok 4.3 (Non-reasoning)Granite 4.0 H 1B
Intelligence Index31.08.0
Coding Index25.12.7
Math Index6.3
Output speed (tok/s)143.70.0
Blended price ($/1M)$1.56$0.00
Time to first token (s)0.42s0.00s
aime
aime 256.3%
artificial analysis coding index25.102.70
artificial analysis intelligence index31.008.00
artificial analysis math index6.30
gpqa65.8%26.3%
hle6.5%5.0%
ifbench47.6%26.2%
lcr24.7%6.3%
livecodebench11.5%
math 500
mmlu pro27.7%
scicode37.4%8.2%
tau265.8%19.6%
terminalbench hard18.9%0.0%

Benchmark data from Artificial Analysis.