deepseek
Catalog updated on 2026-09-07. Current recommended models are listed below. See each model page for supported inputs, outputs, context and pricing. Current recommended models: DeepSeek-V4-Flash-0731, DeepSeek-V4-Pro-0813
- deepseek-v4-flash: Input text; Output text - deepseek-v4-pro: Input text; Output text
High-value reasoning tasks; code generation and completion; long-document analysis; Chinese-language understanding; agent workflows; batch data processing; academic research and education.
Catalog registration does not grant model access. Check channel routing, account permissions, quotas and API compatibility before production use. Official reference: https://api-docs.deepseek.com/quick_start/pricing
| Model | Input / 1M | Output / 1M | Cache Read / 1M | Per 1M |
|---|---|---|---|---|
| 1M | ||||
| 1M | ||||
| - | 1M | |||
| 1M | ||||
| - | 1M | |||
| 1M |