# Exa MCP

> Exa MCP finds what you actually need on the web using contextual search. It goes way beyond basic keyword matching, understanding the intent behind your questions to surface high-quality, curated results and extract clean text from multiple sources automatically.

## Overview
- **Category:** ai-frontier
- **Price:** Free
- **Tags:** neural-search, semantic-similarity, web-data, contextual-search, content-curation, ai-search

## Description

This connector lets your AI agent look up information across the entire internet like a highly specialized researcher—but without you having to copy and paste 50 URLs. Instead of relying on simple keyword matching, it uses advanced semantic search to understand the *meaning* behind your request. You can ask complex questions about niche topics or compare concepts between different fields, getting answers based on contextual understanding. It also lets you find articles that are similar in topic, not just keywords, and then pull out clean, readable text from those pages for analysis. By connecting this MCP through Vinkius, your agent gets access to powerful web data discovery right where you're building your workflows. You can use it to analyze complex source material or research an entire domain without ever leaving your workspace.

## Tools

### answer
Gets an AI-generated answer based on web content context.

### find_similar
Finds pages that are semantically similar to a given URL, useful for competitive analysis.

### find_similar_with_contents
Identifies pages similar to a source and includes the actual text content for review.

### get_contents
Extracts clean, readable text and metadata from specified web page IDs or URLs.

### search_domain
Performs a search limited to a specific website's domain, ideal for documentation lookups.

### search_keyword
Executes traditional keyword-based searches across the web.

### search_neural
Conducts conceptual or research-topic searches that understand intent using neural embeddings.

### search_recent
Searches for content published recently, useful for news and trending topics.

### search
Performs a general web search returning titles, URLs, and relevance scores.

### search_with_contents
Searches the web while simultaneously extracting page content for review.

## Prompt Examples

**Prompt:** 
```
Neural search for 'the latest breakthroughs in room-temperature superconductors'.
```

**Response:** 
```
I've performed a neural search for you. I found 10 high-quality links including recent pre-prints from arXiv, specialized science blogs, and official university press releases. Which source would you like me to extract the content from?
```

**Prompt:** 
```
Find 5 links similar to 'https://openai.com/blog/instruction-following'.
```

**Response:** 
```
Fetching similar content... I've identified 5 semantically related articles, including papers on RLHF from Anthropic, Google DeepMind's blog on fine-tuning, and technical breakdowns from HuggingFace. Shall I retrieve the metadata for these?
```

**Prompt:** 
```
Extract the cleaned text content from 'https://arxiv.org/abs/2312.00752'.
```

**Response:** 
```
Extraction complete! I've retrieved the cleaned text content for the arXiv paper (ID: 2312.00752). It contains the abstract, full body text, and author metadata. Would you like me to summarize the key findings for you?
```

## Capabilities

### Perform conceptual searches
Run advanced neural searches that understand the meaning and intent of a query, not just matching specific words.

### Analyze content similarity
Find web pages or articles that are semantically related to a specific URL you provide for comparative research.

### Extract raw text from multiple sources
Fetch cleaned, usable text and metadata simultaneously from several URLs in one single request.

### Search within defined domains
Restrict your search to a specific website or knowledge base for focused research or documentation lookups.

### Retrieve web data and metadata
Get structured information about search results, helping you identify the most authoritative sources.

## Use Cases

### Mapping out competitive research gaps
A content strategist wants to see what other blogs are writing about a niche topic. They use find_similar_with_contents on their own best-performing article, and the agent returns five semantically related competitor articles along with the full text for immediate analysis.

### Synthesizing complex academic findings
A researcher needs to know about 'room temperature superconductors.' Instead of running ten separate Google searches, they run a neural search and receive a curated set of high-quality sources (pre-prints, university blogs) ready for content extraction.

### Drafting technical documentation updates
A developer needs to update an internal guide on a specific API. They use search_domain against the official vendor site to pull all necessary code snippets and parameter definitions, ensuring accuracy before writing anything.

## Benefits

- Stop relying on simple keyword hits. Use search_neural to ask your agent complex conceptual questions and get answers based on the web's deep meaning.
- Don't waste time copying text from multiple tabs. The get_contents tool pulls clean, readable content and metadata from several URLs in one go.
- Need to analyze a competitor? find_similar is better than simple searches because it finds pages that are conceptually related, not just keyword-matched.
- Want to keep your research focused? Use search_domain to limit the scope of your investigation to a single website's documentation or site section.
- Improve your prompts automatically. The MCP includes tools like 'Autoprompt' to optimize your queries for the highest accuracy before you hit send.

## How It Works

The bottom line is you get powerful, context-aware web data retrieval without having to write complex API calls yourself.

1. Subscribe to this MCP and provide your Exa API Key in your client settings.
2. Your AI agent sends a natural language request—for example, 'Find articles similar to X.'
3. The MCP executes the appropriate search or extraction tool, returning structured, clean web content directly back to your agent for use.

## Frequently Asked Questions

**How does Exa MCP search differently than standard Google searches?**
Exa uses neural embeddings to understand the meaning of your query, not just the words. It finds results based on context and intent, making it ideal for conceptual research topics.

**Can I use Exa MCP to find content from a specific website?**
Yes. You can restrict searches using search_domain to focus only on one site's documentation or knowledge base, ignoring the rest of the web.

**What is the best tool for finding related articles?**
For competitive analysis, use find_similar. If you need both similar pages and their actual text content, use find_similar_with_contents.

**Does Exa MCP only give me links, or does it extract the text?**
It does both. You can run get_contents to pull clean text from a URL, and search_with_contents will perform the web search while extracting the page content simultaneously.

**Is Exa MCP useful for coding documentation lookups?**
Absolutely. Use search_domain to target specific developer sites or internal knowledge bases, ensuring you only pull relevant technical details.