Provides automated daily synchronization with upstream repository changes while preserving Heroku-specific customizations through GitHub Actions workflows
Allows deployment of the MCP server to Heroku platform with automatic scaling and integration with Heroku Managed Inference and Agents for AI model execution
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Exa MCP Server search for latest AI developments in autonomous vehicles"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Connect directly to Exa's hosted MCP server (instead of running it locally).
https://mcp.exa.ai/mcp?exaApiKey=your-exa-api-keyReplace your-api-key-here with your actual Exa API key from dashboard.exa.ai/api-keys.
Add this to your Claude Desktop configuration file:
{
"mcpServers": {
"exa": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.exa.ai/mcp?exaApiKey=your-exa-api-key"
]
}
}
}npm install -g exa-mcp-serverclaude mcp add exa -e EXA_API_KEY=YOUR_API_KEY -- npx -y exa-mcp-serverTo install the Exa MCP server for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install exa --client claudeRelated MCP server: Exa MCP Server
You can find claude_desktop_config.json inside the settings of Claude Desktop app:
Open the Claude Desktop app and enable Developer Mode from the top-left menu bar.
Once enabled, open Settings (also from the top-left menu bar) and navigate to the Developer Option, where you'll find the Edit Config button. Clicking it will open the claude_desktop_config.json file, allowing you to make the necessary edits.
OR (if you want to open claude_desktop_config.json from terminal)
Open your Claude Desktop configuration:
code ~/Library/Application\ Support/Claude/claude_desktop_config.jsonOpen your Claude Desktop configuration:
code %APPDATA%\Claude\claude_desktop_config.json{
"mcpServers": {
"exa": {
"command": "npx",
"args": ["-y", "exa-mcp-server"],
"env": {
"EXA_API_KEY": "your-api-key-here"
}
}
}
}Replace your-api-key-here with your actual Exa API key from dashboard.exa.ai/api-keys.
The Exa MCP server includes the following tools, which can be enabled by adding the --tools:
web_search_exa: Performs real-time web searches with optimized results and content extraction.
company_research: Comprehensive company research tool that crawls company websites to gather detailed information about businesses.
crawling: Extracts content from specific URLs, useful for reading articles, PDFs, or any web page when you have the exact URL.
linkedin_search: Search LinkedIn for companies and people using Exa AI. Simply include company names, person names, or specific LinkedIn URLs in your query.
deep_researcher_start: Start a smart AI researcher for complex questions. The AI will search the web, read many sources, and think deeply about your question to create a detailed research report.
deep_researcher_check: Check if your research is ready and get the results. Use this after starting a research task to see if it's done and get your comprehensive report.
You can choose which tools to enable by adding the --tools parameter to your Claude Desktop configuration:
{
"mcpServers": {
"exa": {
"command": "npx",
"args": [
"-y",
"exa-mcp-server",
"--tools=web_search_exa,company_research,crawling,linkedin_search,deep_researcher_start,deep_researcher_check"
],
"env": {
"EXA_API_KEY": "your-api-key-here"
}
}
}
}For enabling multiple tools, use a comma-separated list:
{
"mcpServers": {
"exa": {
"command": "npx",
"args": [
"-y",
"exa-mcp-server",
"--tools=web_search_exa,company_research,crawling,linkedin_search,deep_researcher_start,deep_researcher_check"
],
"env": {
"EXA_API_KEY": "your-api-key-here"
}
}
}
}If you don't specify any tools, all tools enabled by default will be used.
For the changes to take effect:
Completely quit Claude Desktop (not just close the window)
Start Claude Desktop again
Look for the icon to verify the Exa server is connected
If you prefer to run the server directly, you can use npx:
# Run with all tools enabled by default
npx exa-mcp-server
# Enable specific tools only
npx exa-mcp-server --tools=web_search_exa
# Enable multiple tools
npx exa-mcp-server --tools=web_search_exa,company_research
# List all available tools
npx exa-mcp-server --list-toolsServer Not Found
Verify the npm link is correctly set up
Check Claude Desktop configuration syntax (json file)
API Key Issues
Confirm your EXA_API_KEY is valid
Check the EXA_API_KEY is correctly set in the Claude Desktop config
Verify no spaces or quotes around the API key
Connection Issues
Restart Claude Desktop completely
Check Claude Desktop logs:
Built with ❤️ by team Exa
You can deploy this MCP server to Heroku with one click using the Heroku Button:
If you prefer to deploy manually:
Clone this repository:
git clone https://github.com/dsouza-anush/exa-mcp-server-heroku.git
cd exa-mcp-server-herokuCreate a new Heroku app:
heroku createSet your Exa API key as a config variable:
heroku config:set EXA_API_KEY=your-api-key-hereOptionally, specify which tools to enable:
heroku config:set ENABLED_TOOLS=web_search_exa,company_research_exa,crawling_exaDeploy to Heroku:
git push heroku mainThis MCP server is fully compatible with Heroku Managed Inference and Agents. To use it:
Deploy the MCP server to Heroku using the steps above.
Attach the MCP server to a Heroku Managed Inference and Agents chat model:
# Replace APP_NAME with your Heroku app name
# Replace MODEL_NAME with your desired model name
heroku ai:models:create MODEL_NAME -a APP_NAME --as INFERENCEYour MCP server will be automatically registered with Heroku Inference and its tools will be available via the /v1/agents/heroku endpoint.
This server uses the following Procfile configuration for Heroku Inference:
web: npm run serve
mcp-search-exa: node ./stdio-server.jsThe mcp-search-exa process is registered with Heroku Inference, following the naming convention required by Heroku (process names must start with "mcp"). The direct Node.js execution ensures proper stdio handling for the MCP protocol.
Both web and mcp-search-exa dynos are configured to scale to 0 by default in the app.json, following Heroku's recommendation for MCP servers. When you attach this app to a Heroku Inference model, Heroku will automatically scale the MCP process as needed for tool execution.
To use your Heroku-deployed MCP server with Claude Desktop:
Add this to your Claude Desktop configuration file:
{
"mcpServers": {
"exa_heroku": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://your-app-name.herokuapp.com/mcp"
],
"env": {
"EXA_API_KEY": "your-api-key-here"
}
}
}
}Replace your-app-name with your Heroku app name and your-api-key-here with your Exa API key.
claude mcp add exa_heroku -e EXA_API_KEY=YOUR_API_KEY -- npx -y mcp-remote https://your-app-name.herokuapp.com/mcpReplace your-app-name with your Heroku app name and YOUR_API_KEY with your Exa API key.
This repository automatically updates daily from the original exa-mcp-server repository while preserving all Heroku-specific customizations. Updates are handled through GitHub Actions.
A GitHub Actions workflow runs daily to:
Fetch the latest changes from the original exa-mcp-server repository
Preserve all Heroku-specific customizations (app.json, Procfile, etc.)
Commit and push the updated code
You can also trigger an update manually by:
Going to the "Actions" tab in the GitHub repository
Selecting the "Auto Update from Upstream" workflow
Clicking "Run workflow"
If you prefer to update manually, you can use the included update script:
Clone your forked repository:
git clone https://github.com/your-username/exa-mcp-server-heroku.git
cd exa-mcp-server-herokuRun the update script:
./update-from-upstream.shReview the changes, then commit and push to your repository:
git commit -am "Updated from upstream with Heroku customizations"
git push origin mainIf you have already deployed to Heroku, update your Heroku app:
git push heroku mainThe update process preserves all Heroku-specific customizations, including:
Heroku button configuration (app.json)
Process definitions (Procfile)
Environment variable handling
Heroku-specific scripts
Deployment instructions
Search and get relevant context for any programming task. Exa-code has the highest quality and freshest context for libraries, SDKs, and APIs. Use this tool for ANY question or task for related to programming. RULE: when the user's query contains exa-code or anything related to code, you MUST use this tool.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Search query to find relevant context for APIs, Libraries, and SDKs. For example, 'React useState hook examples', 'Python pandas dataframe filtering', 'Express.js middleware', 'Next js partial prerendering configuration' | |
| tokensNum | No | Number of tokens to return (1000-50000). Default is 5000 tokens. Adjust this value based on how much context you need - use lower values for focused queries and higher values for comprehensive documentation. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already indicate read-only, idempotent, and non-destructive behavior. The description adds value by emphasizing 'highest quality and freshest context' and the programming domain focus, but doesn't disclose additional behavioral traits like rate limits, authentication needs, or response format details. No contradiction with annotations exists.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is front-loaded with the core purpose and usage rule, but includes some redundancy (e.g., repeating 'exa-code' emphasis). Sentences are generally purposeful, though the 'RULE' phrasing could be more integrated. Overall efficient but with minor verbosity.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity (2 parameters, no output schema), annotations cover safety aspects, and the description provides clear purpose and usage rules. However, it lacks details on response structure or error handling, which would enhance completeness for a search tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, providing full documentation for both parameters. The description doesn't add meaningful parameter semantics beyond what's in the schema, such as explaining query formulation strategies or token usage trade-offs. Baseline score of 3 is appropriate given the comprehensive schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'Search and get relevant context for any programming task' with specific focus on 'libraries, SDKs, and APIs.' It distinguishes from the sibling tool 'web_search_exa' by specifying programming-related content, though it doesn't explicitly contrast their differences beyond domain focus.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides explicit usage guidance: 'Use this tool for ANY question or task related to programming' and includes a mandatory rule: 'when the user's query contains exa-code or anything related to code, you MUST use this tool.' This clearly defines when to use it versus alternatives, though it doesn't specify when NOT to use it for non-programming queries.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Search the web using Exa AI - performs real-time web searches and can scrape content from specific URLs. Supports configurable result counts and returns the content from the most relevant websites.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Websearch query | |
| numResults | No | Number of search results to return (default: 8) | |
| livecrawl | No | Live crawl mode - 'fallback': use live crawling as backup if cached content unavailable, 'preferred': prioritize live crawling (default: 'fallback') | |
| type | No | Search type - 'auto': balanced search (default), 'fast': quick results, 'deep': comprehensive search | |
| contextMaxCharacters | No | Maximum characters for context string optimized for LLMs (default: 10000) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false, covering safety and idempotency. The description adds valuable behavioral context beyond annotations: it mentions real-time web searches, scraping from specific URLs, configurable result counts, and returning content from relevant websites. This provides useful operational details without contradicting annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is appropriately sized at two sentences, front-loading the core purpose. Every sentence adds value: the first defines the tool's function, and the second elaborates on features and output. There's no wasted text, though it could be slightly more structured for optimal clarity.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity, rich annotations (covering safety and idempotency), and 100% schema coverage, the description is reasonably complete. It explains the tool's function and key features. The lack of an output schema is a minor gap, but the description mentions return content, partially compensating. For a read-only search tool, this provides adequate context.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema fully documents all 5 parameters. The description adds minimal parameter semantics beyond the schema, mentioning only 'configurable result counts' (referencing numResults) and 'content from the most relevant websites' (hinting at query relevance). Since the schema does the heavy lifting, the baseline score of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'Search the web using Exa AI - performs real-time web searches and can scrape content from specific URLs.' It specifies the verb (search/scrape) and resource (web/URLs), making the function unambiguous. However, it doesn't explicitly differentiate from its sibling 'get_code_context_exa' beyond mentioning general web search capabilities.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for web searches and content scraping, but provides no explicit guidance on when to use this tool versus its sibling 'get_code_context_exa' or other alternatives. It mentions configurable result counts and relevance, which suggests some context, but lacks clear when/when-not directives or named alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v1.0.0company_research_exacrawling_exadeep_researcher_checkdeep_researcher_startget_code_context_exalinkedin_search_exaweb_search_exa5 fields changedInput schema / properties / contextMaxCharactersAdded value: +{
+ "description": "Maximum characters for context string optimized for LLMs (default: 10000)",
+ "type": "number"
+}Input schema / properties / livecrawlAdded value: +{
+ "description": "Live crawl mode - 'fallback': use live crawling as backup if cached content unavailable, 'preferred': prioritize live crawling (default: 'fallback')",
+ "enum": [
+ "fallback",
+ "preferred"
+ ],
+ "type": "string"
+}Input schema / properties / numResults / descriptionPrevious value: -"Number of search results to return (default: 5)"New value: +"Number of search results to return (default: 8)"Input schema / properties / query / descriptionPrevious value: -"Search query"New value: +"Websearch query"Input schema / properties / typeAdded value: +{
+ "description": "Search type - 'auto': balanced search (default), 'fast': quick results, 'deep': comprehensive search",
+ "enum": [
+ "auto",
+ "fast",
+ "deep"
+ ],
+ "type": "string"
+}company_research_exacrawling_exadeep_researcher_checkdeep_researcher_startlinkedin_search_exaweb_search_exaScored across 2 tools
The two tools have clearly distinct purposes: get_code_context_exa is specialized for programming-related searches with high-quality, fresh context for code, libraries, SDKs, and APIs, while web_search_exa is a general web search tool for real-time information and URL scraping. There is no overlap in functionality, making it easy for an agent to select the correct tool based on the query content.
Both tools follow a consistent naming pattern: they use snake_case and start with a verb (get, search) followed by a noun (code_context, web), with a suffix (_exa) to indicate the server. This uniformity makes the tool names predictable and easy to understand, enhancing usability.
With only 2 tools, the server feels thin for its apparent scope of providing search capabilities via Exa AI. While the tools cover distinct areas (code and general web), a typical search server might benefit from additional tools for more granular operations (e.g., filtering, advanced queries) or other domains, making this count borderline too few for robust functionality.
The server covers basic search needs with tools for code-specific and general web searches, but there are notable gaps. For example, it lacks tools for updating or managing search results, handling different content types beyond code and web pages, or providing configuration options beyond result counts. This limits the surface for more complex agent workflows, though core search functionality is present.
Real-time web search, reasoning, and research through Perplexity's API
Web search, fetch, extract, and research for AI agents. Markdown output + AI-synthesized answers.
Web search and page-reading for AI agents. One-click OAuth connect, or a Caesar API key.
Enable AI assistants to perform web searches using Perplexity's Sonar Pro.
