Exa-pool-mcp: Lightweight MCP Server for AI Search Integration

Exa Pool MCP

Exa Pool MCP is a lightweight MCP server that encapsulates the Exa Pool API into a toolset that can be called by AI assistants, facilitating users to conduct searches and content retrieval through tools like Claude.

Search tools Developer tools #Search tool #API encapsulation #AI assistant .Python

rating : 2.5 points

downloads : 7.0K

update time : 2026-01-13

Open Site

What is Exa Pool MCP?

Exa Pool MCP is a bridge connecting AI assistants and the Exa search engine. It allows AI assistants like Claude to directly call the functions of the Exa search engine, including searching web content and finding relevant information, thereby enhancing the AI's answering ability and information accuracy.

How to use Exa Pool MCP?

You only need to complete simple installation and configuration to enable your AI assistant to gain web search capabilities. After configuration, the AI assistant will automatically use the Exa search engine to find the latest information when needed.

Use cases

When you need an AI assistant to answer questions about the latest events, technical documents, news, or any questions that require real - time information, Exa Pool MCP enables the AI assistant to search for information on the Internet like a human being and provide more accurate and up - to - date answers.

Main features

Network search integration

Enable AI assistants to access the Exa search engine and find network information in real - time

Lightweight design

A simple and efficient MCP server implementation with low resource consumption and fast response speed

Simple configuration

You can quickly enable it by simply setting the API key and base URL

AI assistant compatibility

Designed specifically for AI assistants like Claude and seamlessly integrated into the conversation process

Advantages

Free to use: Based on the Exa Pool project, it provides a relatively free usage quota

Easy to integrate: Simple configuration steps can enable AI to gain search capabilities

Real - time information: Enable AI to access the latest network content

Open - source and transparent: The code is open - source, and users can view and verify its working principle

Limitations

Usage limit: The free version has a limit on the number of uses

Dependent on external services: Requires the normal operation of the Exa Pool API service

Configuration requirements: Users need a certain technical foundation for configuration

Relatively basic functions: The functions may be simplified compared to the official Exa API

How to use

Download the file

Download the exa_pool_mcp.py file to your local directory

Get API credentials

Get the API key and base URL from the Exa Pool project

Configure the MCP server

Add the MCP server configuration in Claude Code

Restart the application

Restart Claude Code to apply the configuration changes

Usage examples

Technical research

When you need to understand the latest technological developments or programming frameworks, the AI assistant can automatically search for relevant documents and tutorials

News and information

Get detailed information about current hot news or events

Learning assistance

Find learning materials or explain complex concepts

Frequently Asked Questions

Is Exa Pool MCP free?

Do I need programming knowledge to use it?

Is this tool safe?

Which AI assistants are supported?

How is the search speed?

Related resources

Exa Pool GitHub

Source code and documentation of the Exa Pool project

Exa official documentation

Official documentation and API reference of the Exa search engine

Model Context Protocol

Official documentation and specifications of the MCP protocol

FastMCP project

A framework for quickly building MCP servers

Exa official API

Exa official API service, suitable for high usage limits or commercial use

exa_search

Search the web using Exa's AI - driven search engine. This tool uses neural embeddings for intelligent web search and returns relevant results and their metadata.

Parameters

query : str*

Description

Search query (required)

Parameters

num_results : int*

Description

Number of results to return (1 - 100, default: 10)

Parameters

search_type : str*

Description

Search algorithm - 'auto' (default), 'neural', 'fast' or 'deep'

Parameters

include_text : bool*

Description

Whether to include page text content in the results (default: False)

exa_get_contents

Get clean, parsed content from one or more web pages. Retrieve complete page content through automatic caching and real - time crawling fallback mechanisms.

Parameters

urls : List[str]*

Description

List of URLs to get content from (required, up to 100)

Parameters

include_text : bool*

Description

Include cleaned text content (default: True)

Parameters

include_html : bool*

Description

Include original HTML content (default: False)

exa_find_similar

Find web pages similar to a given URL using semantic similarity. Use Exa's embedding technology to find pages with similar meanings and content.

Parameters

url : str*

Description

Reference URL to find similar pages (required)

Parameters

num_results : int*

Description

Number of similar pages to return (1 - 100, default: 10)

Parameters

include_text : bool*

Description

Whether to include page text content (default: False)

exa_answer

Get AI - generated answers using Exa's Answer API. Perform a search and use the LLM to generate direct answers with references.

Parameters

query : str*

Description

Question to answer (required)

Parameters

include_text : bool*

Description

Whether to include the full text of the reference source (default: False)

exa_create_research

Create an asynchronous in - depth research task. Start an in - depth web research task, explore the web, collect sources, and return structured results with references. This is an asynchronous operation.

Parameters

instructions : str*

Description

Research task description (required, up to 4096 characters)

Parameters

model : str*

Description

Research model - 'exa - research - fast', 'exa - research' (default) or 'exa - research - pro'

exa_get_research

Get the status and results of a research task. Retrieve the current status and results (if completed) of a previously created research task.

Parameters

research_id : str*