← All comparisons

gpt-oss-20B (high) vs Claude 4.1 Opus (Non-reasoning)

OpenAI vs Anthropic — side-by-side benchmark comparison

gpt-oss-20B (high)Claude 4.1 Opus (Non-reasoning)
Intelligence Index24.536.0
Coding Index18.5
Math Index89.3
Output speed (tok/s)267.044.7
Blended price ($/1M)$0.09$32.81
Time to first token (s)0.42s1.63s
aime
aime 2589.3%
artificial analysis coding index18.50
artificial analysis intelligence index24.5036.00
artificial analysis math index89.30
gpqa68.8%
hle9.8%
ifbench65.1%
lcr30.7%
livecodebench77.7%
math 500
mmlu pro74.8%
scicode34.4%
tau260.2%
terminalbench hard10.6%

Benchmark data from Artificial Analysis.