← All comparisons

Gemma 3 4B Instruct vs Claude 4.1 Opus (Reasoning)

Google vs Anthropic — side-by-side benchmark comparison

Gemma 3 4B InstructClaude 4.1 Opus (Reasoning)
Intelligence Index6.342.0
Coding Index2.936.5
Math Index12.780.3
Output speed (tok/s)0.044.5
Blended price ($/1M)$0.05$32.81
Time to first token (s)0.00s8.55s
aime6.3%
aime 2512.7%80.3%
artificial analysis coding index2.9036.50
artificial analysis intelligence index6.3042.00
artificial analysis math index12.7080.30
gpqa29.1%80.9%
hle5.2%11.9%
ifbench28.3%55.4%
lcr5.7%66.3%
livecodebench11.2%65.4%
math 50076.6%
mmlu pro41.7%88.0%
scicode7.3%40.9%
tau25.0%71.4%
terminalbench hard0.8%34.3%

Benchmark data from Artificial Analysis.