DeepSeek
Operated by DeepSeek · Web-connected AI platform
DeepSeek provides web search capabilities for retrieving current information from the web. DeepSeek documents Internet Search in its user-facing products and a server-side web search tool in its API. In the official documentation reviewed by ELNIQ GEO, we did not find a dedicated crawler identity or User-Agent that publishers can use to identify DeepSeek's web retrieval traffic or manage it specifically through robots.txt.
How DeepSeek accesses the web
DeepSeek documents Internet Search as a way to provide real-time answers using information from the web. Its current API also includes a built-in server-side Web Search tool that models can invoke when a request requires current external information.
This establishes web retrieval as a documented DeepSeek capability. However, in the official DeepSeek documentation reviewed by ELNIQ GEO, we did not find a publisher-facing crawler identity, dedicated User-Agent, published IP range, or verification method for identifying those retrieval requests at the website level.
What this means for your site
- DeepSeek can use web search to retrieve current information for AI-generated answers.
- ELNIQ GEO did not find a documented DeepSeek-specific User-Agent or robots.txt token for selectively controlling this web access.
- Without a documented publisher-facing bot identity, individual server requests should not be classified as DeepSeek traffic solely from assumed request characteristics.
- User referrals from DeepSeek and server-side retrieval by DeepSeek are separate traffic signals.
Identification & monitoring
Based on the official DeepSeek documentation reviewed by ELNIQ GEO, we did not find a dedicated publisher-facing User-Agent, bot identity, published IP range, or verification method for DeepSeek's web search traffic.
ELNIQ GEO therefore does not list a DeepSeek-specific User-Agent or robots.txt rule based on the official documentation currently reviewed. This profile should be updated if DeepSeek publishes a dedicated crawler identity or publisher control mechanism.