Microsoft announced Web IQ, a set of grounding APIs that let AI agents pull 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 through a task.

What Web IQ Does

Web IQ uses a rebuilt retrieval stack based on the Bing index, redesigning how content is indexed, ranked, and selected. AI agents search repeatedly under tight time constraints across multiple steps.

The API returns passages and ‘structured evidence objects’ instead of full web pages, so AI models receive only the useful parts of a page. Every token an AI model processes costs money and adds latency, so fewer tokens with better information means cheaper and faster answers.

Microsoft summarizes it as “fewer tokens in, better answers out, lower cost per call.”

Performance Claims

Microsoft uses GDSAT (grounding satisfaction) to measure if info is fresh and trustworthy. They claim Web IQ scores higher than competitors based on 3,000 sample queries.

The company reports sub-165ms response times at P95, nearly 2.5 times faster than competitors, based on tests across five data centers.

On token efficiency, Microsoft reports that Web IQ delivers the same quality with fewer tokens as the volume of results increases.

Publisher Controls

Web IQ follows the same robots 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’s open-sourced embedding model to find relevant content and additional models to rank and select passages.

According to the announcement, these models are trained for how they’ll be used inside AI reasoning, not for standalone benchmark scores.

For fast search at scale, the system extends DiskANN, Microsoft’s technology for searching large indexes without loading everything into memory.

Why This Matters

Microsoft has been building toward this. Bing Webmaster Tools added AI citation data in February, mapped grounding queries to cited pages in March, and previewed Citation Share at SEO Week. Those tools show publishers how AI systems use their content, and Web IQ is the other side of that. It’s what AI systems would use to pull the content in the first place.

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 may not overlap, as Microsoft’s own grounding framework post described earlier this year.

Looking Ahead

Web IQ is accepting expressions of interest but hasn’t announced general availability, pricing, or which AI platforms will use it. Microsoft hasn’t said whether existing Copilot or Bing Chat grounding uses Web IQ, or whether Web IQ is separate from those systems.