Skip to main content

Web Search APIs, priced per call

Choose the level of intelligence your workload needs.

Start building now

Free

View MCP config docs
  • No API key, no account, no credit card
  • 100 Web Search API queries/day via MCP
  • Sign up only when you need more

Production workloads

Pay as you go

Compare pricing
  • No minimums, no seats, no platform fee
  • All endpoints, SDKs and MCP
  • Auto top-up so jobs never stall

Enterprise teams

Custom

Talk to our humans
  • Zero Data Retention, SOC 2, DPA-ready
  • Custom QPS and dedicated capacity
  • SLA-backed support

Powering web search for leading enterprises

APIs and features

All built on the same real-time index. Pick based on how much you want to build yourself.

Get an API key

Web Search

You get the sources and do the reasoning.

Endpoint

Latency

Price

Real-time web and news results, optimized for programmatic access in AI applications.

Web Search APIFree

Live search through the MCP server, no API key—up to 100 queries a day. Prototype or drop live search into an agent before committing.

Latency

p50 ~0.4s

Price

Free

Web Search APIStandard

Full endpoint access: Highlights, Knowledge, extraction, and domain filters. Built for production traffic at scale.

Latency

p50 ~0.4s

Price

$5

/1k calls

Fetches any webpage or PDF and parses into clean Markdown or HTML. Use it when you know which page you want and need its content reproduced accurately. The parse_effort parameter sets how much effort to spend parsing PDF pages, and has no effect on web pages:

ExtractLite

Extracts the content embedded in each page. Works well for digitally created PDFs, such as a text-based contract, at the lowest cost.

Latency

1-10s per URL

Price

$0.90

/1k URLs

ExtractStandard

Extracts the content embedded in each page and sends complex pages, such as charts and images, through an advanced model, billed per page. Use it for the most complete output, such as an earnings report with revenue charts. This is the default.

Latency

1-10s per URL

Price

+$0.50

/1k PDF pages

Inference Router

New

Run leading AI models through one API.

Endpoint

Price

Access leading open-weight models through one API, with built-in reliability and lower inference costs. Choose this when you want to run AI workloads without managing multiple model providers.

Price

Varies by model

Explore models & pricing

Agentic Research

We do the reasoning and hand back the answer.

Endpoint

Latency

Price

One call turns a question into a cited answer with the sources it used. Choose this over Web Search when you want the answer written for you, not a list of links to synthesize yourself.

Latency

p50 ~2.7s

Price

$5

/1k calls

Multi-step research on the open web. It plans, searches repeatedly, reads full pages, and cross-references before answering. One research_effort parameter tunes cost and depth across five levels:

ResearchLite

Simple factual lookups in a single pass. No structured output.

Latency

Under 2s

Price

$12

/1k calls

ResearchStandard

Balanced depth for most production research

Latency

10-30s

Price

$50

/1k calls

ResearchDeep

Complex questions that span several sources and need cross-referencing.

Latency

Under 2min

Price

$100

/1k calls

ResearchExhaustive

High-stakes questions where claims get verified across dozens of sources.

Latency

Under 5min

Price

$450

/1k calls

ResearchFrontier

Research too long to finish inside a request. Background mode only.

Latency

~5min

Price

$1,200

/1k calls

The Research engine tuned for financial questions, at the same price and latency. Use it for company, market, and competitive analysis that needs grounded numbers.

Finance ResearchDeep

Earnings summaries, competitive benchmarking, and multi-quarter trends.

Latency

Under 2 min

Price

$110

/1k calls

Finance ResearchExhaustive

Full 10-K analysis, due diligence, and cross-market research.

Latency

Under 5 min

Price

$500

/1k calls

Decide what to build, then find API.

Pick what you're building. We'll point you to the endpoint that fits, what it costs, and when to reach for something else.

Web Search

Web Search API

$5

/1k calls

Ranked web and news results your model reads and reasons over. You keep full control of synthesis — ideal for RAG.

$10 free credit = 2,000 searches

Enterprise-ready

Get custom pricing, higher rate limits, enterprise security, and dedicated support.

Talk to a human
  • ZDR available

    All queries and data can be automatically purged—giving you complete control over what's stored and for how long.

  • SOC 2 certified

    Our infrastructure is independently audited to ensure best-in-class data security, privacy, and availability.

  • DPA-ready

    Designed to seamlessly accommodate Data Processing Agreements.

  • CCPA

    Compliant with California Consumer Privacy Act requirements for data handling and consumer rights.

  • Custom QPS limits

    Enterprise-grade concurrency with performance optimization for high-volume usage available on request.

  • Dedicated support

    Direct access to our engineering team for onboarding, optimization, and priority issue resolution.

  • The reliability and latency improvements when using You.com were decisive factors. We cut P50 web search latency from over five seconds to sub-one second and pushed our success rate to three 9s or higher.

    Jonathan Lucuix-Andre

    Technical Staff, Factory

Frequent questions about API pricing

One request to an endpoint is one billable call, whatever it returns. A Web Search call returning 1 result and one returning 100 cost the same. Extract is billed per URL, plus per PDF page with the default Standard parse effort.