Nano GPT logo
NanoGPT
Provider logo

DeepSeek V3.2 Exp Thinking

deepseek-ai/deepseek-v3.2-exp-thinking
BackTry Model
Provider logo

DeepSeek V3.2 Exp Thinking

deepseek-ai/deepseek-v3.2-exp-thinking
BackTry Model

Deepseek V3.2 Exp Thinking, Deepseek's latest model offering far better performance especially on longer contexts than its predecessors. Current flagship model by Deepseek. FP8.

Model weights

Context Window

163.8K

Max Output

65.5K

Avg output tokens (7d)

865 tokens

39%

Input Price (Auto)

$0.28/1M

Output Price (Auto)

$0.42/1M

Cache Read (Auto)

$0.14/1M

Capabilities

Benchmarks

Performance metrics and benchmarks

Sourced from Artificial Analysis.

Intelligence Index

25.9

Better than 64% of models compared

Coding Index

44.2

Better than 57% of models compared

Reasoning

GPQA Diamond

Graduate-level scientific reasoning

79.7%

Better than 70% of models compared

HLE

Humanity's Last Exam

14.9%

Better than 69% of models compared

IFBench

Instruction-following benchmark

54.1%

Better than 65% of models compared

T²-Bench Telecom

Conversational AI agents in dual-control scenarios

33.9%

Better than 42% of models compared

AA-LCR

Long context reasoning evaluation

72.3%

Better than 84% of models compared

GDPval-AA

Economically valuable tasks

18.3%

CritPt

Research-level physics reasoning

2.9%

Coding

SciCode

Python programming for scientific computing

37.7%

Better than 62% of models compared

Terminal-Bench Hard

Agentic coding and terminal use

31.1%

Better than 73% of models compared

LiveCodeBench

Contamination-free coding benchmark

78.9%

Better than 91% of models compared

Math

AIME 2025

American Invitational Mathematics Examination 2025

87.7%

Better than 85% of models compared

Knowledge

MMLU-Pro

Professional and academic subject knowledge

85.0%

Better than 91% of models compared

AA-Omniscience Accuracy

Proportion of correctly answered questions

33.0%

AA-Omniscience Hallucination Rate

Rate of incorrect answers among non-correct responses

82.7%

Last updated Aug 16, 2026

Artificial Analysis

Providers

Auto routing is available for this model. Explicit provider selection is not available.

Loading provider options…

Compare DeepSeek V3.2 Exp Thinking with similar models from the same provider or model family.

DeepSeek V3.2 Exp

deepseek-ai/deepseek-v3.2-exp

Deepseek V3.2 Exp, Deepseek's latest model offering far better performance especially on longer contexts than its predecessors. Current flagship model by Deepseek. FP8.

DeepSeek V3.1 Terminus

deepseek-ai/DeepSeek-V3.1-Terminus

DeepSeek-V3.1-Terminus. The latest update builds on V3.1's strengths while addressing key user feedback. Language consistency improvements (fewer CN/EN mix-ups, no random chars), stronger Code Agent & Search Agent performance, and more stable, reliable outputs across benchmarks. FP8.

DeepSeek V3.1

deepseek-ai/DeepSeek-V3.1

DeepSeek-V3.1 is a hybrid model that supports both thinking mode and non-thinking mode. It does better at tool calling and agent tasks, and has higher thinking efficiency than its predecessor. This is the non-thinking version. Quantized at FP8.

DeepSeek R1 0528

deepseek-ai/DeepSeek-R1-0528

The new (May 28th) Deepseek R1 model.

DeepSeek V3.1 Terminus (Thinking)

deepseek-ai/DeepSeek-V3.1-Terminus:thinking

Thinking-enabled DeepSeek-V3.1-Terminus with improved language consistency, upgraded Code/Search Agents, and stronger stability and reliability versus V3.1. FP8.

DeepSeek V3.1 Thinking

deepseek-ai/DeepSeek-V3.1:thinking

Thinking enabled version of Deepseek V3.1. DeepSeek-V3.1 is a hybrid model that supports both thinking mode and non-thinking mode. It does better at tool calling and agent tasks, and has higher thinking efficiency than its predecessor. Quantized at FP8.