DeepSeek
AI model APIs with thinking controls, tool calls, and web search.
Flash API off-peak rates
$0.15 input / $0.60 output
Per 1M tokens; uncached input. Peak rates are double.
Pricing
Pricing not verified
- Use compatible API clients to connect DeepSeek models, choose whether they reason before answering, and connect responses to external tools.
- The standalone Search and Fetch APIs give agents and retrieval applications structured search results and readable page content.
- Keep in mind: API workloads needing one flat rate: model charges vary by tokens, cache status, and peak hours; Search and Fetch are billed separately per successful request.
What is DeepSeek?
DeepSeek provides AI model APIs for developers building text and reasoning workflows. Its API supports OpenAI- and Anthropic-compatible clients, thinking and non-thinking modes, and tool calls. Separate Search and Fetch APIs let applications find current web pages and retrieve their contents as Markdown.
Build text and reasoning workflows
- Compatible API formats let existing OpenAI- or Anthropic-compatible clients access DeepSeek after configuration changes.
- Thinking controls switch between thinking and non-thinking modes for different application workflows.
- Tool calls in thinking mode let the model alternate reasoning with external tool use before its final answer.
Add live web content to applications
- Web search returns structured results for a query, giving applications pages to evaluate and use.
- Page retrieval converts a single web page into Markdown for an application to read or process.
User reviews
No reviews yet. Be the first to review this tool.
Log in to write a review.
Join the Aivolut newsletter
Occasional practical AI tips, app updates, and offers. Unsubscribe anytime.