← All comparisons

GPT-5 Codex (high) vs Claude 3.5 Sonnet (Oct '24)

OpenAI vs Anthropic — side-by-side benchmark comparison

GPT-5 Codex (high)Claude 3.5 Sonnet (Oct '24)
Intelligence Index44.615.9
Coding Index38.930.2
Math Index98.7
Output speed (tok/s)173.20.0
Blended price ($/1M)$3.44$6.56
Time to first token (s)7.64s0.00s
aime15.7%
aime 2598.7%
artificial analysis coding index38.9030.20
artificial analysis intelligence index44.6015.90
artificial analysis math index98.70
gpqa83.7%59.9%
hle25.6%3.9%
ifbench74.1%
lcr69.0%
livecodebench84.0%38.1%
math 50077.1%
mmlu pro86.5%77.2%
scicode40.9%36.6%
tau286.8%
terminalbench hard37.9%

Benchmark data from Artificial Analysis.