- Published on
DeepSeek search and fetch – $0.75 per 1000 successful requests
- Authors
- Name
- DeepHorus
DeepSeek has launched two new API endpoints, Search and Fetch, priced at $0.75 per 1000 successful requests. The Search API finds web pages for a query and returns structured results, while the Fetch API retrieves a single web page as Markdown. Both are designed for agents and RAG applications to access live web content.
The move expands DeepSeek's API offerings beyond language model inference. As AI agents increasingly require real-time information, search capabilities have become essential. DeepSeek's in-house search engine powers the Search API, reducing reliance on third-party providers. The company has been steadily building out its developer ecosystem, and this addition fills a critical gap for real-time data access.
Billing is per successful request, not by tokens or number of results. A request is billable only if it returns HTTP 200 and produces a result—Search must return a non-empty results array, and Fetch must return page content. Failed requests, empty search results, and timeouts are not billed. The count parameter controls how many results are returned but does not affect the per-request price. For example, a search that returns zero results is not billed, encouraging developers to experiment.
The APIs use bearer authentication and the base URL https://api.deepseek.com. Search supports query operators like site:, lang:, and after: for refined searches. Responses include fields such as rerank_score, title, url, snippet, summary, content, site_name, site_icon, date, and authority. Fetch returns a page's title and content in Markdown.
Developers building AI agents and RAG applications can now integrate live web search without managing their own infrastructure. At $0.75 per 1000 successful requests, the pricing is competitive. This could accelerate the development of autonomous agents that need up-to-date information. Startups and independent developers, in particular, benefit from the low cost and ease of integration.
Early reactions from developers are positive, citing the simplicity and cost-effectiveness. Analysts note that DeepSeek is challenging established search API providers like Google and Bing. However, some question the coverage and freshness of DeepSeek's in-house search engine. Some also wonder how DeepSeek's search engine handles niche or specialized queries compared to established players.
DeepSeek is expected to expand its tool APIs further, possibly adding more data sources. Integration with DeepSeek's language models could create a seamless agent experience. The company may also introduce higher-tier plans for enterprise use. Future updates might include support for additional languages and more granular filtering options.
This launch fits a broader trend of AI companies offering full-stack solutions for agent development. Search-as-a-service is becoming a key component, with competitors like OpenAI and Anthropic also providing similar tools. DeepSeek's move signals its ambition to be a one-stop platform for AI applications.
