DeepSeek V3.1 (Reasoning)

DeepSeekChina

DeepSeek V3.1 (Reasoning) is a specialized variant of the V3.1 model optimized for complex reasoning tasks. It incorporates an enhanced thinking process to deliver more accurate and logical solutions for problems requiring multi-step analysis.

ReasoningCodingCheap

Input / 1M tokens

$0.59

Output / 1M tokens

$1.69

Supported plans

Benchmark history

Evaluations

TAU2

Measured Jul 10, 2026Source

Score

0.37

Terminalbench Hard

Measured Jul 10, 2026Source

Score

0.25

Lcr

Measured Jul 10, 2026Source

Score

0.53

Ifbench

Measured Jul 10, 2026Source

Score

0.42

Aime 25

Measured Jul 10, 2026Source

Score

0.9

Scicode

Measured Jul 10, 2026Source

Score

0.39

Livecodebench

Measured Jul 10, 2026Source

Score

0.78

Hle

Measured Jul 10, 2026Source

Score

0.13

Gpqa

Measured Jul 10, 2026Source

Score

0.78

Mmlu Pro

Measured Jul 10, 2026Source

Score

0.85

Artificial Analysis Math Index

Measured Jul 10, 2026Source

Score

89.7

Artificial Analysis Intelligence Index

Measured Jul 10, 2026Source

Score

20.7

Tau Banking

Measured Jul 10, 2026Source

Score

0.16

Terminalbench V2 1

Measured Jul 10, 2026Source

Score

0.45

Artificial Analysis Coding Index

Measured Jul 10, 2026Source

Score

43.5

Plan availability

Products and plans that support this model

No products or plans have been linked to this model yet.

User ratings

Loading ratings...

Discussion

Thinking... Make sure you are connected to GitHub server