Story perspectives

Microsoft Launches Web IQ API, Cuts Retrieval Latency 2.5×

6/4/2026

Summary

Story summary

Microsoft announced Web IQ, an AI-native grounding API suite, on June 2, 2026.
Web IQ returns passage-level evidence and runs at sub-165 ms p95 latency, about 2.5 times faster than rivals.
The service is built on a re-architected Bing global index and follows Bing’s robots-exclusion and publisher controls.
Microsoft added the GDSAT metric to assess grounding satisfaction for completeness, freshness and authority.
The API targets agentic workloads that need repeated low-latency retrieval.

History

Historical Context

Google launched its Knowledge Graph on May 16 2012, indexing about 570 million entities and boosting click-through rates by roughly 15 % in the first year; the panel’s success spurred competitors to adopt structured, low-latency answers. Building on that model, Microsoft’s 2026 Web IQ API delivers passage-level evidence within sub-165 ms, cementing knowledge bases as core to modern AI services.

1 of 2

Story summary

Microsoft announced Web IQ, an AI-native grounding API suite, on June 2, 2026.
Web IQ returns passage-level evidence and runs at sub-165 ms p95 latency, about 2.5 times faster than rivals.
The service is built on a re-architected Bing global index and follows Bing’s robots-exclusion and publisher controls.
Microsoft added the GDSAT metric to assess grounding satisfaction for completeness, freshness and authority.
The API targets agentic workloads that need repeated low-latency retrieval.

1 / 2