Value-first reasoning and coding, with a clean OpenAI-compatible API.
Value-first reasoning, coding and low-cost API scaling.
Open weights on flagship models (MIT), hosted API alongside.
Permissive weights make redistribution straightforward. Keep the attribution notices with the model you deploy.
Licence breakdownPrices are live from the catalogue and may change without notice.
| Model | Billing | Input / 1M | Output / 1M | Cached / 1M | Endpoints |
|---|---|---|---|---|---|
| deepseek-flash DeepSeek Flash. Use deepseek-flash; this site routes requests through Liangxi DeepSeek-Flash. | Tokens | $0.1493 | $0.597 | $0.002985 | openai, anthropic |
| DeepSeek-V4-Flash No description published | Tokens | $0.1493 | $0.2985 | $0.002985 | openai, anthropic |
| DeepSeek-V3.2 No description published | Tokens | $0.2985 | $0.4478 | $0.0597 | openai, anthropic |
| DeepSeek-V4-Flash-Vision-Exp No description published | Dynamic | $1.5 | $4.5 | $0.05 | openai, anthropic |
| DeepSeek-V4-Flash-0731 A 284B-parameter model activating 13B parameters for fast, high-throughput reasoning with a million-token context. | Dynamic | $1.5 | $4.5 | $0.05 | openai, anthropic |
| DeepSeek-V4-Pro A 1.6-trillion-parameter model with ultra-long context, strong reasoning, coding and long-document processing. | Tokens | $1.79 | $3.58 | $0.1493 | openai, anthropic |
| DeepSeek-V4-Pro-0813 The official DeepSeek V4 Pro release with stronger agent capabilities than the preview. | Dynamic | $4.5 | $13.5 | $0.15 | openai, anthropic |