
Tesla
io.github.mcp-dirv0.1.0Updated Oct 11, 2026
Tesla Fleet API for AI: send navigation destinations, read battery and location, run car commands.
Overview
Lets an assistant read a Tesla vehicle's battery, range, charging, climate, locks and location, and send navigation destinations or run car commands.
- What it does
- Connects to the official Tesla Fleet API using the user's own Tesla account, with no third-party relay. Tools list the account's vehicles and read state (battery, range, charging, climate, locks, doors, odometer, software), location and active route destination. Write tools send a street address, coordinates or the nearest Supercharger to the car's navigation, lock and unlock, turn climate on or off and set temperature, start or stop charging and set the charge limit, and honk or flash the lights. Reads do not wake a sleeping car unless asked; commands wake it when needed.
- When to use it
- Useful when someone wants to check or act on their own Tesla from an MCP client in plain language: send a destination before driving, check battery and range, pre-heat or cool the cabin, manage charging, or find the car in a parking lot.
- Requirements
- A remote MCP endpoint over HTTP; no local package or runtime. Requires a Tesla account authorized via OAuth, a paired virtual key approved on the phone for each car (commands fail without it), and prepaid credits for pay-per-use billing. Tesla Fleet API regions are North America/Asia-Pacific and Europe.
Installation
In SourceWeft
- Open Tesla in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.
Other MCP clients
Add this to your client's mcpServers config.
{
"mcpServers": {
"tesla-mcp": {
"type": "http",
"url": "https://api.mcp.ai/p_tesla"
}
}
}README
Tesla
Tesla MCP, send a destination to your car and control it in plain language
Talk to your Tesla from Claude, ChatGPT or any MCP client. It runs on the official Tesla Fleet API, with your own Tesla account (OAuth) and no third-party relay in the middle. The highlight is navigation: "send this address to my car" puts the route on the car's screen, from a street address, coordinates or the nearest Supercharger. It also reads battery, range, charging, climate, locks and location, and runs commands: lock and unlock, climate on/off and temperature, start/stop charging and charge limit, honk and flash. Commands are signed with Tesla's Vehicle Command Protocol. Pay per use with prepaid credits. Not affiliated with Tesla.
- 🧭 Send a destination to the car: street address, coordinates or the nearest Supercharger, straight to the car's navigation
- 🔋 Battery, range and charging: state of charge, range in km, charging status, limit and time to full
- 📍 Location: where the car is, heading and speed, and the active route destination
- ❄️ Climate: turn it on or off and set the temperature before you leave
- 🔒 Locks, honk and lights: lock, unlock (with a confirmation step), honk or flash to find it in a parking lot
- 😴 Sleep-aware: reads never wake the car unless you ask; commands wake it only when needed
- 💬 Works with any MCP client: Claude Desktop, ChatGPT, Cursor, VS Code, Cline, Continue
Versão em português · Full docs (PT-BR) · Agent skill
One-click install
Claude (Web and Desktop)
Anthropic unified MCP installation at claude.ai/customize/connectors. The same link works for Claude Web and Claude Desktop (just be logged in):
Manual (if the deeplink does not open): claude.ai/customize/connectors → + → Add custom connector → paste Name Tesla and URL https://api.mcp.ai/p_tesla.
Cursor
➕ Install Tesla in Cursor
VS Code (Copilot Chat)
➕ Install Tesla in VS Code
ChatGPT, Manus, OpenClaw and 40+ other clients
Works with any MCP client that speaks MCP over HTTP. The server URL is always:
Per-client details: INSTALL.md.
Example prompts
11 tools available
Details for each tool: docs/ferramentas.md (PT-BR)
What you can ask
Before connecting, three honest notes
- Commands need the virtual key. After you authorize, open
https://www.tesla.com/_ak/api.mcp.aion your phone (Tesla app installed) and approve the key for each car. Reads work without it; commands do not. - Regions. The Tesla Fleet API serves accounts in North America/Asia-Pacific and Europe. Accounts in other regions (for example Brazil) may not connect.
- Sleeping cars. Reads never wake the car on their own, because waking costs
more and uses battery. Ask for it (
acordar: true) or just send a command, which wakes the car when it has to.
Pricing
Prepaid credit wallet, pay per use. See docs/precos.md (PT-BR).
Privacy & data protection
- Sub-processors: Tesla, Inc., the LLM host you choose (Claude, ChatGPT, Cursor, your own agent). Full list in docs/privacidade-lgpd.md.
- Data returned by the tools is sent to the LLM host you choose, a sub-processor outside our control. We recommend plans with training opt-out.
FAQ
Is this Tesla's official MCP? No. This is an independent MCP that talks to Tesla's official Fleet API with your own Tesla account. It is not affiliated with Tesla and does not imply a partnership. Tesla is a trademark of its owner, cited here nominatively.
Is there a third party between me and my car? No relay service like Tessie or Teslemetry. Calls go from the mcp.ai server straight to the Tesla Fleet API, and commands are signed with Tesla's Vehicle Command Protocol using our app key, which you approve on your phone.
Why does a command fail on my car? Almost always the virtual key is not paired yet. Open https://www.tesla.com/_ak/api.mcp.ai on your phone with the Tesla app installed and approve it for that car. The tool returns this link when it happens.
Where are my tokens stored? Encrypted, and used only to call the Tesla Fleet API. They never show up in a tool response or in logs. You can revoke access anytime in your Tesla account.
How much does it cost? Pay per use with prepaid credits, priced on Tesla's own per-request fees (data, command, wake). Your first calls are free to try it.
Support
License
MIT — see LICENSE. The MCP server at api.mcp.ai/p_tesla is proprietary (hosted); this repo (manifests, docs, skills) is MIT.
Source: README.md at commit b74690e
Tools
0Version history
1- v0.1.0LatestOct 11, 2026
