DeepSeek API pricing
DeepSeek V4.1 Flash and V4 Pro use peak/off-peak billing. The comparison table uses peak rates; off-peak is 50% lower.
DeepSeek models — price per 1M tokens
All prices in USD. Catalog updated 2026-09-23; model pages show the source-specific verification date. Cheapest reference cost first.
| Model | Input /1M | Output /1M | Cost / call* | Context | Type |
|---|---|---|---|---|---|
| DeepSeek-V4.1 Flash Peak rates · off-peak 50% lower · legacy V4 Flash/Vision IDs route here | $0.300 | $1.20 | $0.0015 | 1M | Fast / cheap |
| DeepSeek-V4 Pro V4-Pro-0813 · peak rates · off-peak 50% lower | $1.32 | $3.96 | $0.0053 | 1M | Reasoning |
*Reference cost of one call with 1,000 input + 1,000 output tokens — a neutral yardstick. DeepSeek rows use peak rates; their model pages and the calculator include the 50%-lower off-peak tier. Use the calculator for your real usage. Cheapest row highlighted.
How DeepSeek API pricing works
DeepSeek bills per token: an input price for everything you send and a higher output price for what the model generates, both quoted per million tokens. To estimate a real bill, multiply by your monthly request volume — the cost calculator does this across every model at once.
Before committing to volume, always confirm the current numbers on the official DeepSeek pricing page — the list prices here are verified periodically but the market moves fast.