← All comparisons

Claude 4.5 Sonnet (Reasoning) vs DeepSeek V3 0324

Anthropic vs DeepSeek — side-by-side benchmark comparison

Claude 4.5 Sonnet (Reasoning)DeepSeek V3 0324
Intelligence Index43.022.3
Coding Index38.622.0
Math Index88.041.0
Output speed (tok/s)55.00.0
Blended price ($/1M)$6.56$1.21
Time to first token (s)7.02s0.00s
aime52.0%
aime 2588.0%41.0%
artificial analysis coding index38.6022.00
artificial analysis intelligence index43.0022.30
artificial analysis math index88.0041.00
gpqa83.4%65.5%
hle17.3%5.2%
ifbench57.3%41.0%
lcr65.7%41.0%
livecodebench71.4%40.5%
math 50094.2%
mmlu pro87.5%81.9%
scicode44.7%35.8%
tau278.1%47.1%
terminalbench hard35.6%15.2%

Benchmark data from Artificial Analysis.