DeepSeek: DeepSeek V3.2
9.0 / 10
163.8K context
Proprietary
About this model
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...
Benchmark Scores
AIME-2025
93.1
AIME-2026
92.7
LiveCodeBench
83.3
Technical Specs
- Architecture: text->text
- Context Window: 163,840 tokens
- Input Modalities: text
Hardware Requirements
- API-only (no local hardware needed)
Pricing
| Input | Output | Currency |
|---|---|---|
| 0.26899999999999996 / 1M tokens | 0.4 / 1M tokens | USD |