← Models / DeepSeek

DeepSeek V3.2 (Thinking)

671B MoE (37B active)

Same as V3.2 but with thinking/reasoning mode enabled via deepseek-reasoner endpoint.

reasoning OpenAI-compatible Function calling 128K context Released 2025-12

Pricing

Type USD / 1M tokens CNY / 1M tokens
Input $0.28 ¥2
Output $0.42 ¥8
Cached Input $0.028 ¥0.5

vs Western Models

DeepSeek V3.2 (Thinking) vs GPT-5.2 ($1.75/$14)
84% cheaper input 97% cheaper output
DeepSeek V3.2 (Thinking) vs Claude Sonnet 4.6 ($3/$15)
91% cheaper input 97% cheaper output
DeepSeek V3.2 (Thinking) vs Gemini 2.5 Pro ($1.25/$10)
78% cheaper input 96% cheaper output

IDE Setup

Cursor

base_url: https://api.deepseek.com/v1
model: deepseek-reasoner

Cline

base_url: https://api.deepseek.com/v1
model: deepseek-reasoner

Strengths

  • + reasoning
  • + coding
  • + math

Watch Out For

  • ! content censorship
  • ! higher output token usage due to CoT

International Access

Direct access: Yes
Latency from US: 150-300ms
Registration: email
Payment: credit card

More from DeepSeek

Ready to try DeepSeek V3.2 (Thinking)?

Get your API key from the official platform.

Go to DeepSeek →