DeepSeek V3.2 Speciale

Name: DeepSeek V3.2 Speciale
Author: DeepSeek-AI

Retired · 2025-12-15Reasoning modelDeepSeek VDeepSeek V3.2

Release date: 2025-12-01Updated: 2026-09-04Retired: 2025-12-15Views: 1,154

Live demoGitHubHugging Face Compare

Parameters

No data

Context length

128K

Multilingual

Supported

Reasoning ability

4/5

DeepSeek V3.2 Speciale is an AI model published by DeepSeek-AI, released on 2025-12-01, for Reasoning model, and 128K context length, with a 2701.00 score on CodeForces.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

DeepSeek V3.2 Speciale

Model basics

Reasoning traces

Supported

Thinking modes

Thinking Mode (Default)Standard Mode

Context length

128K tokens

Max output length

32K tokens

Model type

Reasoning model

Modality (in / out)

Text → Text

Release date

2025-12-01

Model file size

No data

MoE architecture

Total params / Active params

No data / Not applicable

Knowledge cutoff

No data

DeepSeek V3.2 Speciale

Open source & experience

Code license

DeepSeek License Agreement

Weights license

DeepSeek License Agreement- Commercial use permitted

GitHub repo

N/A

Hugging Face

huggingface.co/deepseek-ai/DeepSeek-V3.2-Specia…

Live demo

N/A

DeepSeek V3.2 Speciale

Official resources

Paper

DataLearnerAI blog

N/A

DeepSeek V3.2 Speciale

API details

API speed

3/5

No public API pricing yet.

DeepSeek V3.2 Speciale

Benchmark Results

DeepSeek V3.2 Speciale currently shows benchmark results led by LiveCodeBench (11 / 250, score 89.60), AIME2025 (25 / 215, score 96), GPQA Diamond (132 / 462, score 87.10). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

General Knowledge

3 evaluations

Benchmark / mode

Score

Rank/total

HLE

Thinking Mode

30.60

197 / 563

HLE

Thinking Mode

28.70

216 / 563

CritPt

Thinking Mode

7.40

85 / 200

General Evaluation

1 evaluations

Benchmark / mode

Score

Rank/total

GPQA Diamond

Thinking Mode

87.10

132 / 462

Coding and Software Engineer

2 evaluations

Benchmark / mode

Score

Rank/total

CodeForces

Thinking Mode

2701

8 / 21

LiveCodeBench

Thinking Mode

89.60

11 / 250

Math and Reasoning

1 evaluations

Benchmark / mode

Score

Rank/total

AIME2025

Thinking Mode

25 / 215

Writing and Creative Capabilities

1 evaluations

Benchmark / mode

Score

Rank/total

Creative Writing

Standard Mode

1273.40

75 / 106

Common Sense Reasoning

1 evaluations

Benchmark / mode

Score

Rank/total

SimpleBench

Standard Mode

52.60

48 / 93

Instruction Following

1 evaluations

Benchmark / mode

Score

Rank/total

IF Bench

Thinking Mode

63.90

108 / 282

Agent Level Benchmark

1 evaluations

Benchmark / mode

Score

Rank/total

Terminal Bench Hard

Thinking ModeTools

34.80

75 / 244

View benchmark analysis Compare with other models

DeepSeek V3.2 Speciale

Publisher

DeepSeek-AI

View publisher details

DeepSeek V3.2 Speciale

Model Overview

DeepSeek V3.2 Speciale is an AI model published by DeepSeek-AI, released on 2025-12-01, for Reasoning model, and 128K context length, with a 2701.00 score on CodeForces.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.