Waze MCP Server
Provides Waze routing and travel-time tools to MCP clients, enabling travel time queries, route alternatives, and address geocoding without an API key.
README
Waze MCP Server
A small Model Context Protocol server that exposes Waze routing and travel-time tools to MCP clients (Claude Desktop, Claude Code, VS Code, Cursor, …).
It uses the free pywaze library, which talks to
Waze's public livemap endpoints directly — the same library Home Assistant's
"Waze Travel Time" integration uses. No API key, no third-party proxy, no cost.
Tools
| Tool | Description |
|---|---|
get_travel_time |
Fastest-route duration + distance between two locations |
get_routes |
Up to N alternative routes, each with street names |
geocode_address |
Resolve an address to latitude/longitude |
origin / destination accept a street address or "lat,lng" coordinates.
Optional params: region (US, NA, EU, IL, AU), vehicle_type (car, taxi, motorcycle),
avoid_toll_roads, avoid_ferries, real_time.
All tools are read-only (advertised via MCP tool annotations) and return structured output (typed JSON). Identical lookups are cached briefly to spare Waze's endpoints (see Configuration).
Requirements
- Python 3.10+ —
pywazeandmcprequire it. (On macOS the system Python 3.9 will not work; install a newer one, e.g. via Homebrew or python.org.)
Setup
python3 -m venv .venv # use a Python 3.10+ interpreter
./.venv/bin/pip install -r requirements.txt
Or install it as a package (uses pyproject.toml), which also adds a waze-mcp
command:
./.venv/bin/pip install -e .
On Windows, the interpreter is
.venv\Scripts\python.exeinstead of.venv/bin/python.
Configuration
All settings are optional and set via environment variables:
| Variable | Default | Purpose |
|---|---|---|
WAZE_MCP_REGION |
EU |
Default region (US/NA/EU/IL/AU) when a call omits it |
WAZE_MCP_TIMEOUT |
60 |
Per-request timeout (seconds) to Waze |
WAZE_MCP_CACHE_TTL |
60 |
Seconds to cache identical lookups (0 disables) |
WAZE_MCP_AUTH_TOKEN |
— | If set, HTTP requests must send Authorization: Bearer <token> |
WAZE_MCP_TRANSPORT |
stdio |
Default transport (stdio/streamable-http/sse) |
WAZE_MCP_HOST / WAZE_MCP_PORT |
127.0.0.1 / 8000 |
HTTP bind address |
CLI flags (--transport, --host, --port, --auth-token) override the
matching environment variable.
Run
./.venv/bin/python server.py
The server speaks MCP over stdio, so it is normally launched by your MCP client rather than run by hand. Use the configs below.
In the client configs below, replace
/path/to/waze-mcpwith the absolute path to your clone.
Run over HTTP (remote hosts)
By default the server uses stdio. To let a remote host connect over HTTP — for
example Microsoft Scout's Remote / Local URL mode — start it with the
streamable-http transport:
./.venv/bin/python server.py --transport streamable-http --host 127.0.0.1 --port 8000
The MCP endpoint is served at http://<host>:<port>/mcp, and an unauthenticated
http://<host>:<port>/health endpoint returns ok for liveness checks.
Optional bearer-token auth — require a token on every request (the /health
endpoint stays open):
WAZE_MCP_AUTH_TOKEN=secret \
./.venv/bin/python server.py --transport streamable-http --host 0.0.0.0 --port 8000
# or pass --auth-token secret
Clients then send Authorization: Bearer secret (e.g. Scout's Bearer token field).
Without a token the HTTP endpoint is unauthenticated. Keep it bound to
127.0.0.1, or set a token and/or place it behind a reverse proxy with TLS when exposing it beyond localhost.
Docker
docker build -t waze-mcp .
docker run -p 8000:8000 waze-mcp # http://localhost:8000/mcp
docker run -p 8000:8000 -e WAZE_MCP_AUTH_TOKEN=secret waze-mcp # with auth
The image runs the streamable-http transport on 0.0.0.0:8000 and ships a
/health healthcheck.
VS Code
Add to .vscode/mcp.json (workspace) or your user mcp.json:
{
"servers": {
"waze": {
"type": "stdio",
"command": "/path/to/waze-mcp/.venv/bin/python",
"args": ["/path/to/waze-mcp/server.py"]
}
}
}
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"waze": {
"command": "/path/to/waze-mcp/.venv/bin/python",
"args": ["/path/to/waze-mcp/server.py"]
}
}
}
Claude Code (CLI)
claude mcp add waze -- \
"/path/to/waze-mcp/.venv/bin/python" \
"/path/to/waze-mcp/server.py"
Notes & caveats
- Unofficial data source. Waze has no official public API.
pywazecalls Waze's internal livemap endpoints — the same ones the website uses. They can change or be rate-limited without notice, and heavy use may be against Waze's Terms of Service. - Regions.
EUcovers Europe; useUS/NAfor North America,ILfor Israel,AUfor Australia. Default isEU. - Privacy. Origin/destination coordinates are sent only to
waze.com. The HTTP client's request logging is quieted toWARNING, so addresses/coordinates are not written to the logs by default.
Development
./.venv/bin/pip install -e ".[dev]" # pytest, pytest-asyncio, ruff
./.venv/bin/ruff check .
./.venv/bin/pytest -q # tests mock Waze — no network needed
See CONTRIBUTING.md. CI runs ruff + pytest on Python 3.10–3.13.
Security
The server is read-only and needs no credentials. For the HTTP transport's auth and hardening notes, see SECURITY.md.
Credits
- Routing/geocoding by
pywaze(Kevin Stillhammer), based on WazeRouteCalculator. - "Powered by Waze."
License
MIT © 2026 Serge Michaux
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.