← All comparisons

Llama 3 Instruct 8B vs Claude 4 Sonnet (Reasoning)

Meta vs Anthropic — side-by-side benchmark comparison

Llama 3 Instruct 8BClaude 4 Sonnet (Reasoning)
Intelligence Index6.438.7
Coding Index4.034.1
Math Index74.3
Output speed (tok/s)83.355.5
Blended price ($/1M)$0.07$6.56
Time to first token (s)0.47s8.92s
aime0.0%77.3%
aime 2574.3%
artificial analysis coding index4.0034.10
artificial analysis intelligence index6.4038.70
artificial analysis math index74.30
gpqa29.6%77.7%
hle5.1%9.6%
ifbench24.6%54.7%
lcr0.0%64.7%
livecodebench9.6%65.5%
math 50049.9%99.1%
mmlu pro40.5%84.2%
scicode11.9%40.0%
tau20.0%64.6%
terminalbench hard0.0%31.1%

Benchmark data from Artificial Analysis.