DeepSeek's language model release DeepSeek-V3.
Pricing
Official site · May 12, 2026, 1:07 PM
Input / 1M
$0.27
Output / 1M
$1.10
Cached input / 1M
$0.07
DeepSeek publishes separate cache-hit and cache-miss input token pricing; inputPricePer1M is recorded as the cache-miss rate.
DeepSeek’s official API pricing page lists DeepSeek-V3 with separate input-token rates for cache miss and cache hit, plus an output-token rate.
Model Intelligence
Context window
128,000 tokens
Arena ranking
32
Benchmarkable
Yes
Model level
release
Intelligence Index
16.5
Coding Index
16.4
Math Index
26
MMLU Pro
0.75
GPQA
0.56
HLE
0.04
LiveCodeBench
0.36
SciCode
0.35
MATH-500
0.89
AIME
0.25
AIME 2025
0.26
IFBench
0.35
LCR
0.29
TerminalBench Hard
0.07
TAU2
0.23
Recent stories
0 linked stories
No linked stories yet.