Alpha Vantage MCP Server

The official Alpha Vantage API MCP server enables LLMs and agentic workflows to seamlessly interact with real-time and historical stock market data through the Model Context Protocol (MCP). Add this server to your favorite apps such as Claude, Claude Code, Cursor, VS Code, and many more to give them access to comprehensive financial data.

Quickstart

To use the server, <a href="https://www.alphavantage.co/support/#api-key" onclick="gtag('event', 'mcp_getKey')">get your free Alpha Vantage API key</a>, copy it to your clipboard, then follow the instructions below for the agentic tool/platform of your interest.

⚡ We use an optimization technique called Progressive Tool Discovery to minimize the number of tokens consumed (hence the cost incurred) while preserving the superb response quality when the MCP server is invoked. You can learn more about this technique here

⭐ View MCP source code on Github

👉 Any questions? Please contact support@alphavantage.co

Connection Examples

Remote Server Connection:

https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY

Local Server Connection:

uvx marketdata-mcp-server YOUR_API_KEY

Setup Instructions by Use Case

💡💡 I'm just exploring

Just exploring what to build? Check out the Trading Agents stock research tool (#1 trending on Github), which is powered by Alpha Vantage market data with no setup required.

💬📊 Power your chatbot with financial data

Requirements:

• Claude Pro account (or higher tier)

Claude Remote Server Connection

To connect Claude (Web or Desktop) to this MCP server:

📺 Watch the setup tutorial - Click the image below to watch a step-by-step video guide:

📺 Already have your Alpha Vantage MCP server set up? Below are a few examples of Claude performing various stock analysis & charting tasks:

Query Param Option (Recommended):

1. Go to claude.ai/settings/connectors (Web) or Settings → Connectors (Desktop)
2. Click "Add Custom Connector"
3. Add the MCP server URL with your API key: https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY (replace YOUR_API_KEY with your actual Alpha Vantage API key)
4. Click "Connect"

OAuth Option:

1. Go to claude.ai/settings/connectors (Web) or Settings → Connectors (Desktop)
2. Click "Add Custom Connector"
3. Add the MCP server URL: https://mcp.alphavantage.co/mcp
4. Click "Connect"
5. Enter your Alpha Vantage API token
6. Click "Authorize Access"

Claude Local Server Connection

See Claude Desktop MCP docs for more info.

Install uv (a modern Python package and project manager):

curl -LsSf https://astral.sh/uv/install.sh | sh

Open Claude Desktop developer settings and edit your claude_desktop_config.json file to add the following configuration:

{
  "mcpServers": {
    "alphavantage": {
      "command": "uvx",
      "args": ["marketdata-mcp-server", "YOUR_API_KEY"]
    }
  }
}

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Requirements:

• ChatGPT Plus account (or higher tier)
• Developer mode enabled (beta feature) - see OpenAI's Developer Mode documentation

Setup:

1. Go to ChatGPT Settings → Apps
2. Click on Advanced settings → enable Developer mode
3. Return to the Apps submenu, then click Create app.
4. MCP Server URL: https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY (replace YOUR_API_KEY with your actual Alpha Vantage API key).
5. Authentication: No authentication

Note: When you return to your conversation, your UI will clearly indicate that you are in Developer mode. ChatGPT will retain memory of any previous remarks made within the same chat session; however, these sessions may not be persisted, meaning the conversation generated while Developer mode is active may not be saved or accessible once closed.

Troubleshooting for ChatGPT Plus Accounts

While Developer mode is available for both Plus and Pro accounts, on Plus accounts, first-time MCP tool execution may exhibit inconsistent behavior. This primarily affects large data payloads returned from the MCP server. If you notice that MCP calls abruptly terminate without returning any JSON response, try the following:

1. Confirm Connection

   • Confirm the connection by prompting:
     code

     List 3 functions available in Alpha Vantage MCP server.
     

   • You should see endpoints like TIME_SERIES_DAILY and TIME_SERIES_INTRADAY.

2. Force JSON Passthrough with simple payload

   • Use the following prompts in sequence to prime ChatGPT to surface the full payload:
     code

     Run ADD_TWO_NUMBERS from Alpha Vantage MCP Server using 3 and 6, and then show the JSON response.
     

   • If successful, ChatGPT should ask you to Confirm or Deny the call. Click Confirm.
   • If it doesn't, use the 🔄 "Try Again..." button and see if you get the Confirm/Deny prompt.
   • If this doesn't work, ask ChatGPT the following, and if it starts Thinking make it Skip, so that it doesn't try the call again and instead outputs its conversational response.
     code

     Why aren't you receiving a JSON response?
     

   • Then, regardless of the response, prompt ChatGPT
     code

     Explicitly request a return value from the tool response to verify what the server outputs.
     

   • Typically, this gets ChatGPT to ask you to Confirm or Deny the call. Click Confirm.
   • Sometimes, ChatGPT will insist it has no active connection to Alpha Vantage MCP Server. Just tell it to "try again".
   • If payloads still fail to appear, start a New Chat and try Step #2 again.

3. Force JSON Passthrough with large payload

   • Once step #2 has been successful, repeat step #2 with the larger call:
     code

     Run TIME_SERIES_DAILY using NVDA, and then show the JSON response.
     

   • Use the fallback steps from Step #2 if ChatGPT does not ask you to Confirm or Deny the call.
   • Once you observe ChatGPT successfully return part of the raw JSON payload from the larger call, you can typically continue on with more conversational queries.

🤖📈 Create agentic workflows for quantitative investing

Requirements:

• OpenAI account with access to Agent Builder (e.g., billing details added, organization verified)

📺 Watch the setup tutorial. (Click image below.)

OpenAI Agent Builder Connection

To connect OpenAI Agent Builder to this MCP server:

1. Do not add a separate MCP Server object. Instead, click on your agent in OpenAI Agent Builder to add the MCP Server directly to the agent.
2. Find the Tools section of the agent and click the + button to add a new tool
3. Select MCP Server
4. Click "+ Server" in the upper right of the "Add MCP server" menu
5. Configure the MCP server with the following settings:
   • URL: https://mcp.alphavantage.co/mcp
   • Label: Alpha Vantage MCP Server (or any name you prefer)
   • Description: Financial market data and technical indicators (or any description you prefer)
   • Authentication: Select "Access token / API key" (should be default)
   • Access Token: Enter your Alpha Vantage API key
6. Click Connect
7. If successful, the next prompt will display all the tools. Scroll to the bottom and click Add.

Recommended Agent Instructions:

Add the following instruction to your agent's configuration to optimize performance:

You are a helpful financial agent with access to market data through Alpha Vantage MCP Server.

IMPORTANT: Alpha Vantage functions are accessed via wrapper tools:
- Use TOOL_LIST to see available functions (TIME_SERIES_DAILY, RSI, COMPANY_OVERVIEW, etc.)
- Use TOOL_CALL with the format: TOOL_CALL(tool_name="FUNCTION_NAME", arguments={...})
- Example: TOOL_CALL(tool_name="TIME_SERIES_DAILY", arguments={"symbol": "AAPL", "outputsize": "compact"})

To use the Alpha Vantage MCP server with OpenAI Agents SDK, see our example agent that demonstrates:

• Interactive financial analysis agent
• Session management for conversation continuity
• Real-time tool execution with Alpha Vantage data
• Support for both HTTP and stdio MCP connections

The example includes a complete setup guide and configuration templates.

💻💵 Code up fintech apps

Requirements:

• ChatGPT Plus account (or higher tier)

See OpenAI Codex for more information.

📺 Watch the setup tutorial. (Click image below.) This video additionally provides guidance on handling setup errors and building complete end-to-end applications which are empowered to fetch market data to power dynamic, data-driven visualizations.

• Download the companion .md file from Github Gist.
• For additional support links, please check the video description on YouTube.

Install uv (a modern Python package and project manager):

curl -LsSf https://astral.sh/uv/install.sh | sh

Install Codex CLI v0.34 or later to avoid compatibility issues.

# Install (if not already installed)
npm install -g @openai/codex

# Or update to the latest version
npm update -g @openai/codex

# Verify installation and version
codex --version

Add the following configuration to your Codex MCP server settings by editing ~/.codex/config.toml:

[mcp_servers.alphavantage]
command = "uvx"
args = ["marketdata-mcp-server", "YOUR_API_KEY"]

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Run codex in your terminal from your project directory. First-time users will be guided through additional prompts.

Then connect with:

/mcp

See VS Code MCP docs for more info.

Create .vscode/mcp.json (your VS Code MCP config file) in your workspace, and paste into it one of the following configurations, depending on whether you’re connecting remotely or running the server locally.

VS Code Remote Server Connection

Paste the following into .vscode/mcp.json:

{
  "servers": {
    "alphavantage": {
      "type": "http",
      "url": "https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY"
    }
  }
}

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

VS Code Local Server Connection

First, install uv (a modern Python package and project manager):

curl -LsSf https://astral.sh/uv/install.sh | sh

Then, paste the following into .vscode/mcp.json:

{
  "servers": {
    "alphavantage": {
      "type": "stdio",
      "command": "uvx",
      "args": ["marketdata-mcp-server", "YOUR_API_KEY"]
    }
  }
}

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Open the Chat view and select Agent mode

See Cursor MCP docs for more information.

Paste one of the following configurations into your Cursor ~/.cursor/mcp.json file, depending on whether you’re connecting remotely or running the server locally. You may also install in a specific project by creating .cursor/mcp.json in your project folder.

Cursor Remote Server Connection

Configure Cursor by editing ~/.cursor/mcp.json:

{
  "mcpServers": {
    "alphavantage": {
      "url": "https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY"
    }
  }
}

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Cursor Local Server Connection

First, install uv (a modern Python package and project manager):

curl -LsSf https://astral.sh/uv/install.sh | sh

Then, paste the following into your Cursor ~/.cursor/mcp.json file:

{
  "mcpServers": {
    "alphavantage": {
      "command": "uvx",
      "args": ["marketdata-mcp-server", "YOUR_API_KEY"]
    }
  }
}

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Requirements:

• Claude Pro account (or higher tier)

See Claude Code MCP docs for more information.

Run one of the following commands, depending on whether you’re connecting remotely or running the server locally.

Claude Code Remote Server Connection

claude mcp add -t http alphavantage https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Claude Code Local Server Connection

First, install uv (a modern Python package and project manager):

curl -LsSf https://astral.sh/uv/install.sh | sh

Then run:

claude mcp add alphavantage -- uvx marketdata-mcp-server YOUR_API_KEY

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Run claude in your terminal from your project directory.

Then connect with:

/mcp

See Gemini CLI Configuration for more information.

Run one of the following commands, depending on whether you’re connecting remotely or running the server locally.

Gemini CLI Remote Server Connection (Recommended):

gemini mcp add -t http alphavantage https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Gemini CLI Local Server Connection:

Install uv (a modern Python package and project manager):

curl -LsSf https://astral.sh/uv/install.sh | sh

Then run:

gemini mcp add alphavantage uvx marketdata-mcp-server YOUR_API_KEY

Replace YOUR_API_KEY with your actual Alpha Vantage API key.

Manual Configuration:

1. Open the Gemini CLI settings file. The location is ~/.gemini/settings.json (where ~ is your home directory).
2. Add the following to the mcpServers object in your settings.json file (replace YOUR_API_KEY with your actual Alpha Vantage API key):

{
  "mcpServers": {
    "alphavantage": {
      "httpUrl": "https://mcp.alphavantage.co/mcp?apikey=YOUR_API_KEY"
    }
  }
}

Or, for a local server (replace YOUR_API_KEY with your actual Alpha Vantage API key):

{
  "mcpServers": {
    "alphavantage": {
      "command": "uvx",
      "args": ["marketdata-mcp-server", "YOUR_API_KEY"]
    }
  }
}

If the mcpServers object does not exist, create it.

Examples

Below are four example prompts and Claude's expected responses when using the MCP server.

1. Fetch and display structured market data across assets

Prompt: “Give me last week’s OHLCV data for IBM”

Expected Behavior:

• Claude will use progressive discovery to find and invoke the correct tool, in this case the TIME_SERIES_DAILY function.
• Claude will make the necessary request to the TIME_SERIES_DAILY function and receive the relevant response.
• Claude will then respond directly in the chat interface with both text and image content:

Here's IBM's OHLCV data for last week (March 16–20, 2026):

IBM had a strong Tuesday rally to $256.11, then pulled back through the rest of the week. Friday saw a sharp selloff to close at the session low of $241.77 on more than double the average weekly volume — likely some notable news or broad market pressure driving that move.

2. Use technical indicators and price data to conduct technical analysis

Prompt: “Get NVDA’s hourly price data for the last two days. Compute RSI as well, and plot both hourly with guidelines at 30 and 70 for RSI”

Expected Behavior:

• Claude will use progressive discovery to find and invoke the correct tools, in this case the TIME_SERIES_INTRADAY function and the RSI function.
• Claude will make two separate requests, one to the TIME_SERIES_INTRADAY function and one to the RSI function. Claude will receive the relevant responses.
• Claude will then make an HTML artifact with the requested visualizations:

3. Visualize fundamental data

Prompt: “Help me visualize AAPL’s earnings surprise trends for the past 4 quarters. Compare actual EPS to analyst estimates”

Expected Behavior:

• Claude will use progressive discovery to find and invoke the correct tool, in this case the EARNINGS function.
• Claude will make the necessary request to the EARNINGS function and receive the relevant response.
• Claude will do any necessary calculations on top of the returned data, and create an HTML artifact:

4. Compare performance across asset classes

Prompt: “Compare gold, silver, and Bitcoin performance over the past 5 years”

Expected Behavior:

• Claude will use progressive discovery to find and invoke the correct tools, in this case the TIME_SERIES_MONTHLY function and the DIGITAL_CURRENCY_MONTHLY function.
• Claude will make three separate requests, one to the TIME_SERIES_INTRADAY function using the GLD symbol, one to the TIME_SERIES_INTRADAY function using the SLV symbol, and one to the DIGITAL_CURRENCY_MONTHLY function using the BTC symbol. Claude will receive the relevant responses.
• Claude will do any necessary calculations on top of the returned data, and create an HTML artifact:

Tools Reference

Note: The tools listed below are accessed through the MCP server's TOOL_CALL wrapper function.

Table of Contents - API Tools

• core_stock_apis
• options_data_apis
• alpha_intelligence
• fundamental_data
• forex
• cryptocurrencies
• commodities
• economic_indicators
• technical_indicators
• ping

💡 Each of these MCP tools maps to a corresponding Alpha Vantage API endpoint. If you are interested in the full API specs (in addition to the brief tool descriptions below), please refer to the Alpha Vantage API documentation.

CORE_STOCK_APIS

OPTIONS_DATA_APIS

ALPHA_INTELLIGENCE

FUNDAMENTAL_DATA

FOREX

CRYPTOCURRENCIES

COMMODITIES

ECONOMIC_INDICATORS

TECHNICAL_INDICATORS

PING