Microsoft announced Web IQa set of core APIs that allow AI agents to extract information from Bing’s search index.
The company calls Web IQ “a search engine for AI systems.” Where Bing helps people find web pages, Web IQ helps AI agents find information they can use while reasoning about a task.
What Web IQ does
Web IQ uses a rebuilt retrieval stack based on the Bing index, rethinking how content is indexed, ranked and selected. AI agents perform repeated searches under tight deadlines and in multiple stages.
The API returns passages and “structured evidence objects” instead of full web pages, so AI models only receive useful parts of a page. Each token processed by an AI model costs money and adds latency. So fewer tokens with better information mean cheaper and faster answers.
Microsoft sums it up like this: “fewer tokens in, better responses, lower cost per call.”
Performance claims
Microsoft uses GDSAT (grounding satisfaction) to measure whether information is recent and trustworthy. They claim that Web IQ scores higher than their competitors based on 3,000 sample queries.
The company reports response times of less than 165ms at P95, nearly 2.5x faster than competitors, based on testing across five data centers.
Regarding the effectiveness of tokens, Microsoft reports that Web IQ provides the same quality with fewer tokens as the volume of results increases.
Editor Controls
Web IQ follows the same bot exclusion rules and publisher preferences that Bing already honors. The company is also working with the IETF and other industry groups on standards for how AI systems access web content.
Technical details
Web IQ uses Microsoft open source integration model to find relevant content and additional templates for categorizing and selecting passages.
According to the announcement, these models are trained for how they will be used in AI reasoning, not for standalone benchmark scores.
For fast searching at scale, the system extends DiskANN, Microsoft’s technology for searching large indexes without loading everything into memory.
Why it matters
Microsoft has worked in this direction. Bing Webmaster Tools adding AI citation data in February, basic queries mapped to cited pages in March, and share quote in preview at SEO week. These tools show publishers how AI systems are using their content, and Web IQ is the other side of that. This is what AI systems would use first to extract content.
In Web IQ, information is returned as passages rather than full pages. What makes a page rank well in traditional search and what makes a passage useful for grounding cannot overlap, as Microsoft does. grounding frame post described earlier this year.
Looking to the future
Web IQ is accepting expressions of interest but has not announced its general availability, pricing, or which AI platforms will use it. Microsoft has not clarified whether the existing grounding of Copilot or Bing Chat uses Web IQ, or whether Web IQ is separate from those systems.
