# DeepSeek-V3.2-Speciale

> DeepSeek's reasoning-maxed, open-weight olympiad model that rivals Gemini 3 Pro

- Maker: DeepSeek
- Line: DeepSeek V3
- Canonical page: https://ai-tldr.dev/models/deepseek-v3-2-speciale/
- Updated: 2026-06-18

## In short

DeepSeek-V3.2-Speciale is the reasoning-maxed sibling of DeepSeek-V3.2, both released by DeepSeek on December 1, 2025 as part of the DeepSeek V3 line. It has a 128K-token context window. You can use it through DeepSeek, HuggingFace and OpenRouter.

## Facts

| Fact | Value | Source |
| --- | --- | --- |
| Context window | 128K (max output 128K) | [1] |
| Input price | $0.28 / 1M tokens (cache miss) (cached $0.028 / 1M tokens (cache hit)) | [2] |
| Output price | $0.42 / 1M tokens | [2] |
| Licence | MIT · open weights | [1] |
| Released | 1 December 2025 | [3] |
| Parameters | 671B total / 37B active (MoE) | [1] |
| Knowledge cutoff | Not disclosed | [1] |

## Sources

1. [DeepSeek-V3.2-Speciale on HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Speciale)
2. [DeepSeek-V3.2 Release Announcement](https://api-docs.deepseek.com/news/news251201)
3. [DeepSeek-V3.2-Exp GitHub (DSA architecture)](https://github.com/deepseek-ai/DeepSeek-V3.2-Exp)

## Overview

DeepSeek-V3.2-Speciale is the reasoning-maxed sibling of DeepSeek-V3.2, both released by DeepSeek on December 1, 2025 as part of the DeepSeek V3 line. Where the standard V3.2 balances capability with efficiency and adds tool-use during thinking, Speciale is deliberately tuned for raw reasoning power: it is a "thinking-only" model that does not support tool calls and is designed for deep math, coding, and reasoning problems rather than agentic workflows.

Under the hood, DeepSeek-V3.2-Speciale is a Mixture-of-Experts model with 671 billion total parameters and roughly 37 billion active per token (HuggingFace lists 685B once the Multi-Token Prediction module is counted). It is built on DeepSeek Sparse Attention (DSA) combined with Multi-head Latent Attention, the same long-context architecture that lets the V3.2 family process a 128K-token window cheaply. Speciale was trained exclusively on reasoning data with a reduced length penalty during reinforcement learning, and it borrows the dataset and reward method from DeepSeekMath-V2 to push mathematical-proof ability.

The model is open weight under the MIT license and downloadable from HuggingFace. DeepSeek positioned Speciale as a research and community-evaluation release: at launch it was served through a temporary API endpoint that expired on December 15, 2025, at the same pricing as V3.2. DeepSeek reports that V3.2-Speciale rivals Gemini-3.0-Pro on hard reasoning benchmarks and earned gold-medal-level results at the 2025 IMO, CMO, ICPC World Finals, and IOI.

## Benchmarks

- AIME 2025 (Pass@1): 96% ([source](https://arxiv.org/html/2512.02556v1))
- HMMT Feb 2025 (Pass@1): 99.2% ([source](https://arxiv.org/html/2512.02556v1))
- HMMT Nov 2025 (Pass@1): 94.4% ([source](https://arxiv.org/html/2512.02556v1))
- IMOAnswerBench (Pass@1): 84.5% ([source](https://arxiv.org/html/2512.02556v1))
- LiveCodeBench (Pass@1-COT): 88.7% ([source](https://arxiv.org/html/2512.02556v1))
- GPQA Diamond (Pass@1): 85.7% ([source](https://arxiv.org/html/2512.02556v1))
- Humanity's Last Exam (HLE, Pass@1): 30.6% ([source](https://arxiv.org/html/2512.02556v1))

## Pricing

- Input: $0.28 / 1M tokens (cache miss) per 1M tokens
- Cached input: $0.028 / 1M tokens (cache hit) per 1M tokens
- Output: $0.42 / 1M tokens per 1M tokens

DeepSeek-V3.2-Speciale was served at the same pricing as DeepSeek-V3.2 via a temporary API endpoint that expired December 15, 2025. Weights remain available under MIT for self-hosting.

Source: https://api-docs.deepseek.com/news/news251201

## Strengths

- Gold-medal-level performance on the 2025 IMO, CMO, ICPC World Finals, and IOI competitions
- Top-tier scores on hard math and coding benchmarks (96.0% AIME 2025, 99.2% HMMT Feb 2025, 88.7% LiveCodeBench)
- Open weights under a permissive MIT license, fully downloadable and self-hostable
- DeepSeek Sparse Attention (DSA) delivers efficient 128K long-context reasoning
- MoE design activates only ~37B of 671B parameters per token, keeping inference cost low for the capability level
- Released at the same low price as V3.2 (far cheaper than comparable frontier reasoning models)

## Best for

- Solving competition-grade mathematics and writing formal mathematical proofs
- Hard algorithmic and competitive-programming problems (ICPC/IOI-style)
- Research and academic evaluation of frontier open-weight reasoning models
- Long-context reasoning over large documents and codebases (up to 128K tokens)
- Self-hosted reasoning deployments where open weights and data control matter
- Benchmarking and distillation source for smaller reasoning models

## How to access

- [DeepSeek](https://api-docs.deepseek.com/news/news251201) — model id `deepseek-v3.2-speciale`
- [HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Speciale) — model id `deepseek-ai/DeepSeek-V3.2-Speciale`
- [OpenRouter](https://openrouter.ai/deepseek/deepseek-v3.2-speciale) — model id `deepseek/deepseek-v3.2-speciale`

## DeepSeek V3 versions

- [DeepSeek-V3.2](https://ai-tldr.dev/models/deepseek-v3-2/) (1 December 2025)
- DeepSeek-V3.2-Speciale (1 December 2025) — this page
- [DeepSeek-V3.2-Exp](https://ai-tldr.dev/models/deepseek-v3-2-exp/) (29 September 2025)
- [DeepSeek-V3.1-Terminus](https://ai-tldr.dev/models/deepseek-v3-1-terminus/) (22 September 2025)
- [DeepSeek-V3.1](https://ai-tldr.dev/models/deepseek-v3-1/) (21 August 2025)
- [DeepSeek-V3-0324](https://ai-tldr.dev/models/deepseek-v3-0324/) (24 March 2025)
- [DeepSeek-V3](https://ai-tldr.dev/models/deepseek-v3/) (26 December 2024)
- [DeepSeek-V2.5](https://ai-tldr.dev/models/deepseek-v2-5/) (5 September 2024)
- [DeepSeek-V2](https://ai-tldr.dev/models/deepseek-v2/) (May 2024)

## FAQ

### What is DeepSeek-V3.2-Speciale?

DeepSeek-V3.2-Speciale is a reasoning-maxed, open-weight variant of DeepSeek-V3.2, released by DeepSeek on December 1, 2025. It is a 671B-parameter Mixture-of-Experts model (about 37B active per token) built on DeepSeek Sparse Attention. It is a "thinking-only" model with no tool-call support, tuned for deep math, coding, and reasoning tasks.

### How does Speciale differ from the standard DeepSeek-V3.2?

Standard V3.2 balances capability and efficiency and integrates tool-use directly into its thinking mode. Speciale is a higher-compute variant trained exclusively on reasoning data with a reduced length penalty during reinforcement learning. It produces longer reasoning chains, scores higher on hard benchmarks, does not support tool calls, and was released as a temporary research endpoint.

### Is DeepSeek-V3.2-Speciale open source, and what does it cost?

The model weights are open and released under the permissive MIT license, downloadable from HuggingFace for self-hosting. On DeepSeek's API it was served at the same price as V3.2: about $0.28 per million input tokens (cache miss), $0.028 per million on a cache hit, and $0.42 per million output tokens, through a temporary endpoint that expired December 15, 2025.

### How does DeepSeek-V3.2-Speciale perform on benchmarks?

DeepSeek reports it rivals Gemini-3.0-Pro on difficult reasoning workloads. Per the tech report it scores 96.0% on AIME 2025, 99.2% on HMMT Feb 2025, 88.7% on LiveCodeBench, and 85.7% on GPQA Diamond, and it achieved gold-medal-level results at the 2025 IMO, CMO, ICPC World Finals, and IOI.

## Links

- [DeepSeek-V3.2 Release Announcement](https://api-docs.deepseek.com/news/news251201)
- [DeepSeek-V3.2 Tech Report (arXiv)](https://arxiv.org/html/2512.02556v1)
- [DeepSeek-V3.2-Speciale on HuggingFace](https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Speciale)
- [DeepSeek-V3.2-Exp GitHub (DSA architecture)](https://github.com/deepseek-ai/DeepSeek-V3.2-Exp)
- [DeepSeek-V3.2-Speciale on OpenRouter](https://openrouter.ai/deepseek/deepseek-v3.2-speciale)
- [DeepSeek-V3.2-Speciale on LLM-Stats](https://llm-stats.com/models/deepseek-v3.2-speciale)

Source: AI/TLDR LLM registry, https://ai-tldr.dev/models/deepseek-v3-2-speciale/