#19
DeepSeek V3.2 Thinking
deepseek-reasoner-v3.2
0.0
Helpfulness
Instruction Following
Comprehension
Empathy
Creative Writing
Helpfulness
0.0
Instruction Following
0.0
Comprehension
0.0
Empathy
0.0
Creative Writing
0.0
Speed
Avg 45 tok/s
Release Date
September 29, 2025
Lab
DeepSeek
Type
Open Source
Context Size
160K
Max Output Tokens
32.8K
Cost per 1 million tokens
$0.14 / $0.28
Model Inputs*
Text
Model Outputs*
Text
Tool Calling*
Disabled
Overall Assistant Score
An average score combining the 5 main categories.
84.48 pts
Rank #19
63th Percentile
0
Novice
33
Capable
66
Proficient
100
Expert
DeepSeek V3.2 Exp Thinking is a specialized reasoning variant of the V3.2 architecture, designed for complex logical deduction and chain-of-thought processing. It achieves exceptional scores in comprehension and instruction following by utilizing an extended 'thinking' process before generating answers. While less suited for creative or empathetic tasks, it stands as a powerhouse for math, coding, and deep analytical problems.

Intelligence

Overall Score; Higher is better

Kimi K2.5
ChatGPT 4o
Gemini 2.5 …
Claude Haik…
Grok 4 Fast…
DeepSeek V3…
Gemini 3 Fl…
GPT-5 Mini
o4-mini
Grok 4 Fast

Speed

Output Tokens per Second; Higher is better

Gemini 2.5 …
Claude Opus…
Gemini 3 Pro
o3
Claude Opus…
Claude Opus…
DeepSeek V3…
Kimi K2 Thi…
DeepSeek V3…
DeepSeek R1

Price

USD per 1M Tokens; Lower is better

DeepSeek V3…
GLM 4.7 Fla…
Gemini 2.0 …
Gemini 2.0 …
Gemini 2.5 …
DeepSeek V3…
Llama 4 Sco…
Grok 4.1 Fa…
Grok 4.1 Fa…
Grok 4 Fast…

deepseek Models

Overall Score; Same provider comparison

DeepSeek V3…
DeepSeek V3…
DeepSeek V3…
DeepSeek R1
DeepSeek V3…
DeepSeek V3

DeepSeek Family

Overall Score; Same model family comparison

DeepSeek V3…
DeepSeek V3…
DeepSeek V3…
DeepSeek R1
DeepSeek V3…
DeepSeek V3

Closest Rivals

Overall Score; Nearest by overall rank

Kimi K2.5
ChatGPT 4o
Gemini 2.5 …
Claude Haik…
Grok 4 Fast…
DeepSeek V3…
Gemini 3 Fl…
GPT-5 Mini
o4-mini
Grok 4 Fast