Introducing Parallel Decisions API: the first decision engine with world knowledge built in. Learn more.[Learn more.]
Microsoft Azure AI Foundry logo

## What you can do

  • - Give Foundry agents live web search and page extraction inside their reasoning loop.
  • - Return current, source-backed answers with citations from agents built on Foundry Agent Service.

## How it works

The Parallel Search MCP server is a remote MCP tool in the Azure AI Foundry tool catalog, exposing web_search and web_fetch backed by Parallel's Search and Extract APIs.

Attach it to an agent as an MCP tool and the agent calls Parallel mid-loop, receiving token-efficient sourced excerpts; anonymous requests run in fast mode, and a Parallel API key raises rate limits for production use.

## Set up in three steps

  1. 1. Open the Parallel Search MCP listing in the Azure AI Foundry tool catalog.
  2. 2. Attach the MCP server to your agent, passing a Parallel API key as a Bearer token for production rate limits.
  3. 3. Run the agent and verify web_search and web_fetch return current, cited web results.
Add Parallel Search MCP[Add Parallel Search MCP]