← All comparisons

Llama 3.3 Instruct 70B vs Grok 4.3 (low)

Meta vs xAI — side-by-side benchmark comparison

Llama 3.3 Instruct 70BGrok 4.3 (low)
Intelligence Index14.543.9
Coding Index10.731.6
Math Index7.7
Output speed (tok/s)88.1118.3
Blended price ($/1M)$0.62$1.56
Time to first token (s)0.59s6.35s
aime30.0%
aime 257.7%
artificial analysis coding index10.7031.60
artificial analysis intelligence index14.5043.90
artificial analysis math index7.70
gpqa49.8%84.3%
hle4.0%17.3%
ifbench47.1%81.0%
lcr15.0%64.0%
livecodebench28.8%
math 50077.3%
mmlu pro71.3%
scicode26.0%41.9%
tau226.6%88.9%
terminalbench hard3.0%26.5%

Benchmark data from Artificial Analysis.