DeepSeek-V3.2

DeepSeek-V3.2

deepseek-ai/DeepSeek-V3.2

About DeepSeek-V3.2

DeepSeek-V3.2 is a model that harmonizes high computational efficiency with superior reasoning and agent performance. Its approach is built upon three key technical breakthroughs: DeepSeek Sparse Attention (DSA), an efficient attention mechanism that substantially reduces computational complexity while preserving model performance, specifically optimized for long-context scenarios; a Scalable Reinforcement Learning Framework, which enables performance comparable to GPT-5 and reasoning proficiency on par with Gemini-3.0-Pro in its high-compute variant; and a Large-Scale Agentic Task Synthesis Pipeline to integrate reasoning into tool-use scenarios, improving compliance and generalization in complex interactive environments. The model has achieved gold-medal performance in the 2025 International Mathematical Olympiad (IMO) and International Olympiad in Informatics (IOI)

Available Serverless

Run queries immediately, pay only for usage

$

0.27

/

$

0.42

Per 1M Tokens (input/output)

Metadata

Create on

Dec 4, 2025

License

MIT

Provider

DeepSeek

HuggingFace

Specification

State

Available

Architecture

Calibrated

No

Mixture of Experts

No

Total Parameters

671B

Activated Parameters

671B

Reasoning

No

Precision

FP8

Context length

164K

Max Tokens

164K

Supported Functionality

Serverless

Supported

Serverless LoRA

Not supported

Fine-tuning

Not supported

Embeddings

Not supported

Rerankers

Not supported

Support image input

Not supported

JSON Mode

Supported

Structured Outputs

Not supported

Tools

Supported

Fim Completion

Not supported

Chat Prefix Completion

Supported

SiliconFlow Service

Comprehensive solutions to deploy and scale your AI applications with maximum flexibility

60%

lower latency

2x

higher throughput

65%

cost savings

Model FAQs: Usage, Deployment

Learn how to use, fine-tune, and deploy this model with ease.

Ready to accelerate your AI development?

Ready to accelerate your AI development?

Ready to accelerate your AI development?