← All comparisons

Claude 4.5 Sonnet (Reasoning) vs DeepSeek R1 0528 (May '25)

Anthropic vs DeepSeek — side-by-side benchmark comparison

Claude 4.5 Sonnet (Reasoning)DeepSeek R1 0528 (May '25)
Intelligence Index43.027.1
Coding Index38.624.0
Math Index88.076.0
Output speed (tok/s)55.00.0
Blended price ($/1M)$6.56$2.06
Time to first token (s)7.02s0.00s
aime89.3%
aime 2588.0%76.0%
artificial analysis coding index38.6024.00
artificial analysis intelligence index43.0027.10
artificial analysis math index88.0076.00
gpqa83.4%81.3%
hle17.3%14.9%
ifbench57.3%39.6%
lcr65.7%54.7%
livecodebench71.4%77.0%
math 50098.3%
mmlu pro87.5%84.9%
scicode44.7%40.3%
tau278.1%36.5%
terminalbench hard35.6%15.9%

Benchmark data from Artificial Analysis.