DeepSeek V3.2 Speciale
671B MoE (37B active)
High-compute variant of V3.2 optimized for maximum reasoning and agentic performance. No function calling support. Available via third-party providers (OpenRouter).
reasoning OpenAI-compatible 164K context Released 2025-12
Pricing
| Type | USD / 1M tokens |
|---|---|
| Input | $0.4 |
| Output | $1.2 |
vs Western Models
DeepSeek V3.2 Speciale vs GPT-5.2 ($1.75/$14)
77% cheaper input 91% cheaper output
DeepSeek V3.2 Speciale vs Claude Sonnet 4.6 ($3/$15)
87% cheaper input 92% cheaper output
DeepSeek V3.2 Speciale vs Gemini 2.5 Pro ($1.25/$10)
68% cheaper input 88% cheaper output
Strengths
- + reasoning
- + agentic tasks
- + maximum performance variant
Watch Out For
- ! no tool calls
- ! higher cost than base V3.2
International Access
Direct access: Yes
Latency from US: 150-300ms
Registration: email
Payment: credit card
More from DeepSeek
Ready to try DeepSeek V3.2 Speciale?
Get your API key from the official platform.
Go to DeepSeek →