Microsoft Corp. introduced Web IQ, a new suite of application programming interfaces designed to allow artificial intelligence agents to access and utilize information directly from Bing’s search index, aiming to improve reasoning and efficiency.
The global initiative positions Web IQ as a specialized search engine tailored for AI systems, providing targeted passages and structured evidence objects rather than complete web pages, according to Microsoft.
Microsoft stated that the system employs a rebuilt retrieval stack based on the Bing index, which has been optimized for content indexing, ranking and selection specifically for AI agent consumption.
The company claims Web IQ achieves higher grounding satisfaction and significantly faster response times, with 95 percent of responses delivered in under 165 milliseconds, surpassing competitor performance.
Web IQ aims to enhance token efficiency for AI models, which Microsoft said would result in “fewer tokens in, better answers out, lower cost per call.”
The new technology respects existing robots exclusion rules and publisher preferences, according to Microsoft. The company also reported it is collaborating with industry groups, including the IETF, on developing web content access standards for AI.
Microsoft indicated that Web IQ leverages its open-sourced embedding model and additional ranking models. It also extends DiskANN for fast, scalable search capabilities.
While Microsoft is currently accepting expressions of interest for Web IQ, no general availability, pricing details or specific AI platform integrations have been announced.
Source: Search Engine Journal
