DeepSeek V3.2
Provider：DeepSeek
DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
Key specs
- Context：163.8K
- Max output：65.5K
- Tokenizer：DeepSeek
- Released：2025-12-01
Token pricing
- Input price：$0.269 / 1M tokens
- Output price：$0.400 / 1M tokens
- Cache read：$0.135 / 1M tokens
- Blended price：$0.302 / 1M tokens
Modalities
Text
Use Cases
- Deep Reasoning & Analysis：Math proofs, logical reasoning, multi-step planning — ideal for research and financial analysis.（≈ $0.0032/call）
When to pick this model
Signals derived from public pricing and spec fields — not a benchmark. Validate with official docs and your own eval.
- Low input + output price: solid fit for high-frequency calls and batch processing.
- Long max output (>=30K tokens): strong fit for deep report / essay / document-drafting workflows.
FAQ
- What is the token pricing of DeepSeek V3.2? Input price $0.269 / 1M tokens, output price $0.400 / 1M tokens, blended about $0.302 / 1M tokens. Refer to DeepSeek’s official page for the exact rate.
- What context window does DeepSeek V3.2 support? Context window is 163.8K, max output about 65.5K, suitable for long documents, multi-turn chat and complex reasoning.
- What scenarios is DeepSeek V3.2 best for? Based on its capability and pricing, DeepSeek V3.2 fits best: Deep Reasoning & Analysis. See the Use Cases section for details and cost estimates.
- What input/output modalities does DeepSeek V3.2 support? Supports Text modality.
- Is DeepSeek V3.2 a free model? DeepSeek V3.2 is billed per token, not a free model.
- Which provider offers DeepSeek V3.2? DeepSeek V3.2 is offered by DeepSeek.
More models from DeepSeek
Full info：DeepSeek V3.2