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...
| Provider | deepseek |
|---|---|
| Input price | $0.260 / 1M tokens |
| Output price | $0.380 / 1M tokens |
| Context window | 164k tokens |
| Max output | 164k tokens |
| Modalities | text |
| Model ID | deepseek/deepseek-v3.2 |
A 1,000-token prompt with a 500-token answer on DeepSeek: DeepSeek V3.2 costs roughly $0.00045. Use the OpenRouter cost calculator to price your own prompt across 400+ models, or the token counter to measure prompt length first.
PromNest stores your prompt library and lets you run any prompt on DeepSeek: DeepSeek V3.2 with your own OpenRouter key — including side-by-side comparison with other models. Browse 1,000+ ready-made prompt templates to start.