# exerciseapi [Health: Active]

**Category:** 🏃 Sports  
**Repository:** https://github.com/westvegh/exerciseapi-mcp-server  
**GitHub Stars:** 3  
**npm Downloads (last month):** 114  
**Views:** 1  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/exerciseapi

## Description
Give your AI coding agent direct access to 2,198+ exercises via exerciseapi.dev

## Tools
Capabilities this server exposes over MCP:

- **search_exercises** — Search the library by name, category, muscle, equipment, or difficulty. Returns up to 100 exercises.
- **get_exercise** — Fetch full details for a single exercise by ID — instructions, form tips, safety notes, variations, video.
- **list_categories** — List all 12 exercise categories with counts.
- **list_muscles** — List anatomical muscles available as filters.
- **list_equipment** — List equipment types available as filters.
- **get_stats** — Library-wide stats: total count, breakdown by category, video coverage.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "exerciseapi": {
    "command": "npx",
    "args": ["-y","@exerciseapi/mcp-server"],
    "env": {
      "EXERCISEAPI_KEY": ""
    }
  }
}
```

**Requires environment variables:** `EXERCISEAPI_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What exerciseapi MCP server does

The exerciseapi MCP server exposes the exerciseapi.dev REST API as MCP tools for AI coding agents. It provides access to more than 2,198 exercises spanning 12 categories, including strength, calisthenics, yoga, Pilates, mobility, physical therapy, plyometrics, and stretching. Exercise records can include anatomical muscle information, equipment requirements, difficulty, instructions, form tips, safety notes, variations, and demonstration videos where available.

The service is intended for developers creating fitness, workout, training, rehabilitation, or wellness applications. It gives an agent a way to retrieve library-backed exercise information instead of asking the model to invent movements, muscle groups, or form guidance.

## How it works

The exerciseapi MCP server is a thin wrapper around the exerciseapi.dev REST API. Filtering and searching happen on the API side; the MCP process makes those operations available to the connected agent. The server runs as an MCP process launched with `npx`, or as a globally installed command named `exerciseapi-mcp`.

A typical request might ask an agent to find intermediate chest exercises using dumbbells, retrieve the form tips for selected results, or compare two exercises. The agent can search first and then request complete data for a specific exercise by ID. Search results can include up to 100 exercises.

## Setup and configuration

Create an API key through the exerciseapi.dev dashboard before starting the server. Add the package to the MCP configuration for the client you use, with `npx -y @exerciseapi/mcp-server` as the command and the key supplied through the `EXERCISEAPI_KEY` environment variable. Restart the client after changing its configuration.

`EXERCISEAPI_KEY` is required; the server exits during startup if it is absent. `EXERCISEAPI_BASE_URL` is optional and defaults to `https://api.exerciseapi.dev`. It can point the wrapper at another API endpoint for self-hosting or staging tests. The documented free tier allows 100 requests per day.

## Tools and capabilities

The exerciseapi MCP server provides six tools:

- `search_exercises` searches by name, category, muscle, equipment, or difficulty and returns up to 100 matches.
- `get_exercise` retrieves one exercise by ID, including instructions, form tips, safety notes, variations, and video data when available.
- `list_categories` returns the 12 exercise categories and their counts.
- `list_muscles` lists the anatomical muscles that can be used as filters.
- `list_equipment` lists available equipment filter values.
- `get_stats` returns library totals, category breakdowns, and video coverage.

## Limitations and notes

The exerciseapi MCP server depends on a valid API key and the exerciseapi.dev service. Requests are subject to the account's quota; the documented free tier limit is 100 requests per day and resets at midnight UTC. Demonstration videos cover only a growing subset of exercises, so video data is not guaranteed for every result.

If the agent does not invoke the tools, the README recommends checking the MCP client logs to confirm that the process started and making the request explicitly mention the exerciseapi tools. The first `npx` launch may be slower because the package is downloaded. Direct REST API access is available when MCP is not suitable, using the same API key.

_Full upstream README: https://allmcps.com/mcp/exerciseapi/readme_