← All comparisons

DeepSeek V4 Flash (Reasoning, Max Effort) vs Claude 4.5 Sonnet (Reasoning)

DeepSeek vs Anthropic — side-by-side benchmark comparison

DeepSeek V4 Flash (Reasoning, Max Effort)Claude 4.5 Sonnet (Reasoning)
Intelligence Index46.543.0
Coding Index38.738.6
Math Index88.0
Output speed (tok/s)119.355.0
Blended price ($/1M)$0.17$6.56
Time to first token (s)0.86s7.02s
aime
aime 2588.0%
artificial analysis coding index38.7038.60
artificial analysis intelligence index46.5043.00
artificial analysis math index88.00
gpqa89.4%83.4%
hle32.1%17.3%
ifbench79.2%57.3%
lcr63.0%65.7%
livecodebench71.4%
math 500
mmlu pro87.5%
scicode44.9%44.7%
tau295.0%78.1%
terminalbench hard35.6%35.6%

Benchmark data from Artificial Analysis.