DeepSeek V3.2

Non-reasoning

Creator

DeepSeek

Released

2025-12-01

Intelligence

32.1

Artificial Analysis Index

Coding

34.6

Artificial Analysis Index

In $/1M

$0.50

input tokens

Out $/1M

$1.60

output tokens

Blended $/1M

$0.78

3:1 blended

Speed

0

tokens / sec

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
75
Humanity’s Last Exam
11
MMLU-Pro
84
SciCode
39
LiveCodeBench
59
MATH-500
AIME 2025
59
τ²-Bench (agentic)
79
Terminal-Bench Hard
33
IFBench
49
Via Artificial Analysis