Skip to content

Live data for AI agents through MCP or the REST API

Give an agent search results, page content, company profiles and validation checks as tools. Connect the MCP server in minutes, or call the REST API from your own tool layer.

What you end up with
  • Agent answers rest on current sources that can be traced back to a lookup.
  • One server and one balance replace a handful of separate search, scraping and validation integrations.
  • Validation checks sit between the agent and any outbound action, which keeps bad data from turning into bad messages.
[01/Pipeline]

The steps, in order

  1. 1

    Connect the MCP server

    Add https://mcp.1lookup.dev/mcp as an MCP server in Claude, ChatGPT, Cursor or another MCP client. Among its eight tools, run_lookup runs any product by slug. For frameworks without MCP, call POST https://api.1lookup.dev/v1/{product-slug} with a bearer key.

    Your system
  2. 2

    Let the agent find current information

    Give the agent SERP Scraper for live Google results with People Also Ask and related searches, so answers rest on today's web instead of training data.

    SERP Scraper1 cr / search
  3. 3

    Let it read the pages it finds

    Website Scraper turns any URL into clean Markdown and metadata ready for an LLM context window. The agent can open a result, read it and cite it.

    Website Scraper10 cr / page
  4. 4

    Let it look up entities

    For a company named in a conversation, the agent can call Company Profile Lookup with the domain for size, headquarters and funding history, or Google News Search for recent coverage.

    Company Profile Lookup25 cr / match
  5. 5

    Validate before it acts

    Before an agent sends an email or places a call, have it run Email Validation or Phone Validation on the contact. A cheap check at the end of the chain stops many bad actions.

    Email Validation1 cr
[03/FAQ]

Live data for AI agents: common questions

What is the difference between the MCP server and the REST API?

They reach the same products. The MCP server at https://mcp.1lookup.dev/mcp lets an MCP client discover eight tools, including run_lookup for running any product by slug. The REST API suits your own code and agent frameworks that do not use MCP.

How do I control what an agent spends?

Every product has a published price in credits, and all spending comes from one balance. Restrict the agent to the products it needs, set budgets in your own orchestration code, and watch usage in the dashboard.

Are lookups safe to retry?

Requests that fail on our side are never charged, and a repeat of the same input within 7 days is served from cache at no charge. Bulk jobs take an idempotency key, which keeps retried submissions from running twice.

Can an agent act on people data?

Be deliberate about it. Keep people-search products out of agent tool lists unless your use case is allowed, never use results for credit, employment, tenant or insurance decisions, and have a person review actions that affect individuals.

[ Validate ][ Enrich ]
[Get started]

Run your first lookup in two minutes

Create a key, pick a plan and call any product in the catalog. Plans start at $29 a month.

[ Screen ][ Scrape ]