Powering frontier AI labs,
trusted by enterprises

Web IQ gives agents and AI applications up-to-date, comprehensive knowledge about what is changing in the world across web pages, news, images, video, and more.

Get started Read the documentation

Powering top AI platforms

Microsoft Copilot
OpenAI
Mistral

OVERVIEW

Web grounding that brings real-world knowledge to AI agents

Build enterprise-ready agents

Ground agents with permission-aware web knowledge, backed by enterprise safeguards like Zero Query Logging and customizable source controls.

Frontier performance, lower total cost

Get high-quality answers with fast retrieval and ready-to-use context that cuts token consumption and end-to-end costs.

Better knowledge, better outcomes

Give agents access to comprehensive public web information and structured knowledge to help them reason, answer questions, and take action.

FEATURES

Trusted sources. Enterprise-grade controls

Responsible grounding

Web IQ transparently identifies its crawlers and respects publisher preferences expressed via robots.txt. Microsoft is also the only company that gives content owners industry-first, granular permission preferences specific to AI grounding.

Control the sources your agents work with

Web IQ offers source controls to focus grounding on approved websites and exclude specific pages.

See the evidence behind your agents’ work

Review supporting passages, source citations, and metadata to understand where information comes from and trace your agents’ work back to the evidence.

Works with your trusted processes

Generally available in Azure, Web IQ offers familiar procurement, consumption meter and billing. And Azure eligible consumption discounts apply.

Easy deployment, fast production

Connect through REST APIs or MCP to your preferred model and workflow, without rebuilding your stack. And integration with Microsoft IQ makes it easy to orchestrate answers across organizational and external sources.

Meet your enterprise privacy requirements

Zero Query Logging is available for Azure Managed Customers in regulated industries. Learn more.

Engineer working at monitors in a blue-lit office

Give your agents richer information to reason over

Web IQ brings together public web content, licensed content and purpose-built, structured data in one API, so your agents can answer with fresher, more precise evidence.

Public web

Answer with what's happening now. Web pages, news, videos and images, returned with citations.

Licensed content

Use the content you already license. Bring your existing publisher and data-provider licenses. Licensed and web results come back ranked together in one response.

Purpose-built data

Reason over facts, not just scraped pages. Structured, ready-to-use data for places, weather and commerce (private preview), with more on the way.

Web IQ tuned for health

For health questions, Web IQ ranks credible health sources above forums and blogs for better answers. Your agents answer from cited evidence you can trust.

PERFORMANCE

Setting the standard for AI grounding

P95 latency in millisecond (ms) based on 1K queries each from 8 regions across North America, Europe, and Asia. P95 latencies were averaged across 8 regions. The cohort includes each provider's flagship offering that meets a minimum quality threshold (not exhaustive). All queries were fresh so that caching does not affect latencies. Only HTTP responses with 200 code were considered. Latencies were measured with connection pool for all the providers. Analysis was done on 11/8.

Industry-leading speed

Keep complex agent workflows moving with retrieval that's 2× faster at p95 latency than the next-fastest service tested. At just 154 milliseconds, Web IQ minimizes retrieval delays that can compound across multi-step agent workflows.

*Query set consists of 2850 randomly sampled global queries. Curves were plotted by running different settings of web results and the depth of content. The cohort includes providers and their flagship offerings within a quality threshold. Curves truncated on the right to stay within typical token limits and for visualizations. Evals were run between 11/8 and 11/9.

Frontier performance without compromises

High quality with token efficiency determines whether an agent is useful and can scale. Web IQ delivers on both fronts. With multi-step retrieval in a single call and ready-to-use passages, Web IQ reduces retrieval overhead and token consumption.

The open-sourced FreshSearch benchmark evaluates freshness across two dimensions: how effectively grounding providers discover newly published content and how fresh evidence is for time sensitive queries. Scores averaged over sampled runs in November 2026.

Fresh results for up-to-date answers

An agent is only as current as the knowledge and context it’s given. When evaluating how current answers are and the ability to find newly published content, Web IQ shows outstanding ability to return fresh results.

CUSTOMER STORIES

Built for frontier AI labs. Now available to enterprises

The grounding infrastructure behind Copilot and ChatGPT is available for your AI agents and applications. See how organizations across many industries are transforming agent outcomes with Web IQ.

Swipe to explore

PRICING

Simple, consumption-based pricing

The lowest search API price doesn't always mean the lowest cost per answer. What matters is delivering high-quality results efficiently.

Web IQ optimizes for both cost and quality, retrieving relevant, focused context without processing entire webpages. That means fewer tokens, better answers, and lower total cost of ownership.

* Commerce available in private preview
Browse $1 / 1k calls
Web, News, Images, Video, Places, Commerce* $7 / 1k calls
Auto $8 / 1k calls
Auto Advanced $9 / 1k calls
See pricing

FAQ

Frequently asked questions

Microsoft Web IQ is a suite of AI-native web search and grounding APIs that gives LLMs, AI agents and applications access to current information from the web and the world, including web pages, news, images, video, places, weather, and product data. This enables more accurate, up-to-date AI answers with sources users can verify. Web IQ powers leading frontier AI assistants, including Microsoft Copilot and ChatGPT.

Web grounding gives AI access to the latest web information, so it can answer using current facts beyond its training data, with cited sources users can check. It helps most with frequently changing information, like market developments, competitor activity, news, regulatory updates, and more.

Most web search APIs force a tradeoff — fast or accurate, cost-efficient or complete. Web IQ delivers across the board: industry-leading latency and high quality coupled with token efficiency. It's the same platform that AI assistants like Copilot and ChatGPT rely on every day — proven at a global scale, and ready for any AI agent or app to build on.

Web IQ is differentiated by how it accesses content. It takes a responsible approach to web access, bringing together the open web, licensed content, and structured data with enterprise-grade security, compliance, and protections built in.

Web IQ also works with the AI stack you already use. Developers can use their preferred AI model and connect through Model Context Protocol (MCP), without being tied to a single model or platform.

Web IQ powers AI agents and applications that need fresh, external information from the public web. Internal teams can use it for competitive intelligence, monitoring, regulatory research, sales intelligence, and content optimization; customer-facing experiences can use it for research, discovery, recommendations, and specialized agents. It’s best suited for scenarios where information changes frequently, spans many external sources, and needs to be grounded in current, attributable web content.

Yes, Web IQ supports a range of enterprise and regulated use cases. Customers with specialized compliance, governance, or industry requirements can engage with our team to determine the appropriate solution offering.

Web IQ delivers fresher, faster, and more efficient grounding without trade-offs. In our benchmarks, it leads on freshness and speed, with P95 latency of about 154 ms. Web IQ also reaches comparable quality with roughly half the tokens, which lowers your costs, and it keeps improving on quality with our latest retrieval updates.

Web IQ retrieves current information across web pages, news, images, and video from a continuously refreshed index. It also provides structured data for:

- Places: 207.5 million places in 200+ markets, with addresses, phone numbers, opening hours, and coordinates.

- Weather: current conditions, hourly forecasts for the next 24 hours, daily forecasts for 7 days, and government safety warnings in 65 countries.

- Commerce (private preview): more than 40 billion merchant offers.

Web IQ is designed to honor supported publisher controls, including applicable robots.txt directives and other publisher signals recognized by Microsoft.

Web IQ draws from the global web to support AI agents and applications serving different regions and audiences. Language and market coverage can vary, so review supported coverage against the questions your users need to answer.

Yes. Web IQ lets you add web information to the AI model and agent workflow you choose. REST APIs and MCP give you flexibility in how you connect it; check the integration documentation for supported configurations.

Microsoft IQ brings together four integrated layers that power enterprise intelligence across Microsoft. Web IQ brings knowledge from the world outside your organization. Work IQ focuses on work signals, Fabric IQ on business data, and Foundry IQ on organizational knowledge. Together, they help agents understand both your business and the changing world around it.

Create a Web IQ resource in Azure and connect with an API key or Microsoft Entra ID. Call Auto to ground a request across the right sources in one call, or call individual APIs such as Web, News, Images, Videos, Places, or Browse when you want full control. Your agent passes the returned passages to its model to generate a response, and you control how information is selected, used, and cited. Follow the Quickstart or try the Playground to connect your first workflow.

Yes. With bring your own license (BYOL), content you license from a publisher or data provider can be delivered through Web IQ and returned alongside web results in a single ranked response. Your agreement with the provider stays in place, and the provider authorizes delivery through Web IQ. Initial support covers eligible unstructured, long-form content available through URLs. Submit the BYOL form to review your content source and licensing arrangement

Web IQ pricing is usage-based and varies by API - $1 per 1,000 calls for Browse, $7 per 1,000 calls for Web, News, Images, Videos, Places. $8/1,000 calls for Auto, and $9 per 1,000 calls for Auto Advanced. Commerce APIs are available to select groups in private preview for $7 per 1,000 calls. Review the current pricing page for the applicable API rates, then estimate total workflow cost using expected calls per task, query volume, retries, returned content, and model-input tokens. Customer-specific commercial terms may apply.

Web IQ is available to all Azure customers worldwide, except China. It can be used by enterprise companies, small businesses, and individual users. Access is subject to your intended use, Azure account history, and other conditions outlined in our terms of use.

Bring the world to your AI agents

Start building with Web IQ today. Connect your agents to fresh world intelligence and give them the external context they need to answer, reason, and act.

Back to top