Open-Meteo MCP Server
Summary: This server is an MCP weather/climate toolkit that lets an LLM fetch accurate Open-Meteo data — forecasts, 1940-onward history, air quality, marine, flood, seasonal and CMIP6 climate — for any coordinates, with no API key.
Weather forecasts up to 16 days (or recent past via
past_days), with hourly, daily and 15-minute variables, unit and timezone options.Historical weather from 1940 to yesterday via ERA5 reanalysis.
Air quality: PM2.5, PM10, ozone, NO₂, pollen, European/US AQI, UV index.
Marine conditions: wave height, period, direction, swell, sea surface temperature, ocean currents.
Floods: river discharge from GloFAS, optionally all 50 ensemble members.
Seasonal outlooks from weeks to ~7 months ahead, including weekly/monthly anomalies.
Climate projections from CMIP6 models, 1950–2050.
Ensemble forecasts showing uncertainty across members (a model must be chosen).
Model-specific forecasts from DWD ICON, NOAA GFS, Météo-France, ECMWF, JMA, MET Norway and Environment Canada GEM — one model per call, parallel calls to compare.
Location helpers: geocoding of place names/postal codes to coordinates, and elevation lookups (batched).
Solar/radiation variables such as global tilted irradiance using panel tilt and azimuth.
Deployment flexibility: stdio for local clients or stateless HTTP at
/mcpwith API key auth, rate limiting, origin checks, response caching and redirectable endpoints for self-hosted Open-Meteo.
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., "@Open-Meteo MCP Server what's the 7-day weather forecast for Paris?"
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.
Open-Meteo MCP Server
Give any LLM accurate weather data: forecasts, history back to 1940, air quality, marine conditions, floods and climate projections.
Features • Getting started • Tools • Remote deployment • Configuration • Development
A Model Context Protocol server for the free Open-Meteo APIs. Plug it into Claude Desktop, Claude Code or any MCP client, then ask in plain language:
What were the temperatures in London during January 2023?
Compare the ICON and GFS ensemble forecasts for Berlin over the next 5 days.
Give me the current European AQI, UV index and pollen levels in Paris.No API key is needed: Open-Meteo is free for non-commercial use.
Features
17 tools covering forecasts, ERA5 history, air quality, marine, flood, seasonal, ensemble and CMIP6 climate data, plus geocoding and elevation
Model-specific forecasts from DWD ICON, NOAA GFS, Météo-France, ECMWF, JMA, MET Norway and Environment Canada GEM
Built for LLMs: strict input schemas, server instructions that tell the model which tool answers which question, compact JSON responses capped at 25,000 characters
Two transports: stdio for local clients, stateless Streamable HTTP for remote deployments, with API key auth, rate limiting and origin checks
In-memory response cache with per-endpoint TTLs, so repeated questions don't hit Open-Meteo again
Self-hosting friendly: every Open-Meteo endpoint can point at your own instance
Related MCP server: open-meteo-mcp-server
Getting started
You need Node.js 22 or later. Nothing to install beforehand: npx fetches the server on first run.
Claude Desktop
Add the server to your claude_desktop_config.json:
{
"mcpServers": {
"open-meteo": {
"command": "npx",
"args": ["-y", "-p", "open-meteo-mcp-server", "open-meteo-mcp-server"]
}
}
}Claude Code
claude mcp add open-meteo -- npx -y -p open-meteo-mcp-server open-meteo-mcp-serverOther MCP clients
Any client that launches stdio servers works with the same command: npx -y -p open-meteo-mcp-server open-meteo-mcp-server. You can also install it globally with npm install -g open-meteo-mcp-server and run open-meteo-mcp-server.
Every data tool takes coordinates. Ask with a place name and the model will callgeocoding first to resolve it.
Tools
Category | Tool | What it answers |
Core |
| Forecast up to 16 days, picking the best model for the location. Recent past via |
| Historical weather from 1940 to yesterday (ERA5 reanalysis) | |
| PM2.5, PM10, ozone, NO₂, pollen, European and US AQI, UV index | |
| Wave height, period and direction, swell, sea surface temperature | |
| Place name or postal code to coordinates | |
| Terrain height for coordinates | |
Models |
| DWD ICON (Germany, high resolution over Europe) |
| NOAA GFS (global, high resolution over North America) | |
| Météo-France AROME and ARPEGE | |
| ECMWF IFS and AIFS | |
| Japan Meteorological Agency | |
| MET Norway (Nordic countries) | |
| Environment Canada GEM | |
Advanced |
| Forecast uncertainty across ensemble members. |
| Outlook from a few weeks to about 7 months ahead | |
| CMIP6 climate projections, 1950 to 2050 | |
| River discharge from GloFAS |
weather_forecast is the default choice. The model-specific tools are for when a particular model is asked for, or to compare models with one call each.
Responses
Times are GMT unless
timezoneis set.timezone: "auto"uses the location's local time.nullin a series means the model has no value for that time, not zero.Responses over 25,000 characters have their
hourly/daily/minutely_15arrays shortened by the same ratio, keeping series aligned, and gaintruncated: truewith atruncation_message. Narrow the date range or the variables to get everything.
The full list of variables and parameters is in each tool's input schema and in the Open-Meteo documentation.
Remote deployment
Set TRANSPORT=http to serve MCP over Streamable HTTP at /mcp instead of stdio:
TRANSPORT=http HOST=0.0.0.0 PORT=3000 API_KEY=your-secret-key npx open-meteo-mcp-serverClients then send the key with every request, as Authorization: Bearer <key> or X-API-Key: <key>. GET /health answers {"status":"ok"} without a key, for container probes.
The transport is stateless: each POST /mcp is handled on its own, no session ID is issued, and GET / DELETE /mcp answer 405. No tool keeps state between calls, so clients lose nothing.
The server binds to127.0.0.1 by default, so it is reachable only from the local machine. Set HOST=0.0.0.0 to accept remote connections, and set API_KEY whenever you do: without it, the server runs in open mode.
Docker
A prebuilt image is published to the GitHub Container Registry. It already binds to 0.0.0.0:
docker run -d --name open-meteo-mcp -p 3000:3000 \
-e API_KEY=your-secret-key \
ghcr.io/cmer81/open-meteo-mcp:latestTags follow the npm version without the v prefix: latest, 2.5.1, 2.5, 2.
The repository also has docker-compose.yml (prebuilt image) and docker-compose.dev.yml (builds from source). Copy .env.example to .env to configure them.
claude.ai traffic
Every claude.ai user reaches a remote server from Anthropic's outbound range 160.79.104.0/21. That range gets its own rate-limit pool (RATE_LIMIT_ANTHROPIC_RPM) so they don't all share one per-IP budget. Behind a reverse proxy, list the proxy in TRUSTED_PROXIES so the real client IP is seen.
Configuration
All variables are optional.
Server
Variable | Default | Description |
| stdio |
|
|
| HTTP port |
|
| Interface to bind. |
|
| Response cache size, in bytes of serialized JSON. |
The cache keeps forecasts and ensembles for 15 minutes, air quality and marine for 30 minutes, flood for 1 hour, seasonal for 6 hours, archive and climate for 24 hours, geocoding for 7 days and elevation for 30 days. Archive ranges ending within the last 5 days are kept for 1 hour only, since Open-Meteo is still backfilling them. Failed requests are never cached.
The cache counts serialized JSON, but the parsed objects in memory take about 1.2 to 2.6 times as much. A full cache at the default size costs about 50 MB of heap.
HTTP security
Variable | Default | Description |
| unset (open) | Key required on every |
|
| Requests per minute per client IP. IPv6 clients are grouped by /56 |
|
| Requests per minute shared by all claude.ai traffic |
| unset | Comma-separated IPs or CIDRs whose |
| empty | Comma-separated browser origins allowed. Any request with an unlisted |
Custom Open-Meteo instance
Each endpoint can be redirected, for example to a self-hosted Open-Meteo:
Variable | Default |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
In Claude Desktop, pass them through the env key of the server entry.
Skills
The skills/ directory holds two SKILL.md guides that help an assistant pick the right tool and parameters:
Skill | Best for |
Everyday weather: forecasts, history, air quality, marine, elevation | |
Specific models, ensemble uncertainty, seasonal outlooks, climate projections |
For Claude Code, copy them to ~/.claude/skills/:
cp -r skills/open-meteo skills/open-meteo-advanced ~/.claude/skills/For Claude Desktop, upload the relevant SKILL.md into the conversation.
Development
git clone https://github.com/cmer81/open-meteo-mcp.git
cd open-meteo-mcp
npm install
npm run buildCommand | Description |
| Run from source with auto-reload (stdio / HTTP) |
| Unit tests (network mocked) |
| Type checking and Biome linting |
| Calls all 17 tools against the live API through a real MCP client. Needs a prior build |
| LLM-usability benchmark, see below |
To point Claude Desktop at your local build, use "command": "node" with "args": ["/path/to/open-meteo-mcp/dist/index.js"].
Evaluations
evals/evaluation.xml checks whether an LLM given only this server's tools can answer realistic questions. Its 14 questions rely on stable data (ERA5 archive, CMIP6 projections, geocoding, elevation), so the expected answers don't drift.
pip install -r evals/scripts/requirements.txt
export ANTHROPIC_API_KEY=... # or put it in .env
npm run build && npm run eval
npm run eval -- --no-server-instructions # baseline without the server instructionsThe evaluation calls the real Anthropic API for every question and consumes credits. It is a manual check, not part of CI.
Contributions are welcome: see CONTRIBUTING.md.
Available Tools
17 toolsair_qualityAir QualityBRead-onlyIdempotent
Get air quality forecast data including PM2.5, PM10, ozone, nitrogen dioxide, pollen, European/US AQI indices, UV index and other pollutants.
Inputs: latitude (required), longitude (required), hourly, current, +7 more.
climate_projectionClimate ProjectionCRead-onlyIdempotent
Get climate change projections from CMIP6 models for different warming scenarios.
Inputs: end_date (required), latitude (required), longitude (required), start_date (required), +9 more.
dwd_icon_forecastDWD ICON ForecastARead-onlyIdempotent
Get weather forecast from German DWD ICON model. IMPORTANT: Specify exactly one DWD model in the models parameter (e.g., "dwd_icon_global") — only one model per request is supported. For multi-model comparison, make one parallel tool call per model using the appropriate provider-specific tool.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
ecmwf_forecastECMWF ForecastARead-onlyIdempotent
Get weather forecast from ECMWF models via the dedicated /v1/ecmwf endpoint. IMPORTANT: Specify exactly one model in the models parameter — only one model per request is supported. Valid model IDs: "ecmwf_ifs" (IFS HRES 9 km), "ecmwf_ifs04" (IFS open-data 0.4°), "ecmwf_ifs025" (IFS open-data 0.25°, the default), "ecmwf_aifs025" and "ecmwf_aifs025_single" (AIFS machine-learning model, ensemble mean and single run), "ecmwf_ifs_europe_ensemble_mean" and "ecmwf_aifs_europe_ensemble_mean" (Europe-only ensemble means), "best_match". Model IDs with an underscore before the resolution digits (e.g. "ecmwf_ifs_025") are not valid. For multi-model comparison, make one parallel tool call per model using the appropriate provider-specific tool.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
elevationElevationARead-onlyIdempotent
Get elevation data for given coordinates using digital elevation models. Supports batch lookups: pass arrays of the same length for latitude and longitude to query multiple points in one call.
Inputs: latitude (required), longitude (required).
ensemble_forecastEnsemble ForecastBRead-onlyIdempotent
Get ensemble forecasts showing forecast uncertainty with multiple model runs. models is required: the ensemble API has no default model (e.g. "icon_seamless_eps", "ncep_gefs_seamless", "ecmwf_ifs025_ensemble").
Inputs: models (required), latitude (required), longitude (required), tilt, +17 more.
flood_forecastFlood ForecastCRead-onlyIdempotent
Get river discharge and flood forecasts from GloFAS (Global Flood Awareness System).
Inputs: latitude (required), longitude (required), daily, models, +8 more.
gem_forecastGEM ForecastARead-onlyIdempotent
Get weather forecast from Canadian Meteorological Centre (GEM) models. IMPORTANT: Specify exactly one GEM model in the models parameter (e.g., "gem_global" or "gem_regional") — only one model per request is supported. For multi-model comparison, make one parallel tool call per model using the appropriate provider-specific tool.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
geocodingGeocodingARead-onlyIdempotent
Search for locations worldwide by place name or postal code. Returns geographic coordinates (latitude and longitude) and detailed location information. Use this tool when you need to convert a location name (e.g., "Paris", "New York") into precise coordinates (latitude/longitude) that are required by other tools. This is essential when you have a location name but need coordinates for data fetching tools.
Inputs: name (required), count, format, language, +1 more.
gfs_forecastGFS ForecastARead-onlyIdempotent
Get weather forecast from US NOAA GFS model. IMPORTANT: Specify exactly one GFS model in the models parameter (e.g., "ncep_gfs_global") — only one model per request is supported. For multi-model comparison, make one parallel tool call per model using the appropriate provider-specific tool.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
jma_forecastJMA ForecastARead-onlyIdempotent
Get weather forecast from Japan Meteorological Agency (JMA) models. IMPORTANT: Specify exactly one JMA model in the models parameter (e.g., "jma_msm" or "jma_gsm") — only one model per request is supported. For multi-model comparison, make one parallel tool call per model using the appropriate provider-specific tool.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
marine_weatherMarine WeatherCRead-onlyIdempotent
Get marine weather forecast including wave height, wave period, wave direction and sea surface temperature.
Inputs: latitude (required), longitude (required), daily, hourly, +13 more.
meteofrance_forecastMétéo-France ForecastARead-onlyIdempotent
Get weather forecast from French Météo-France models. IMPORTANT: Specify exactly one Météo-France model in the models parameter (e.g., "meteofrance_arome_france" or "meteofrance_arpege_europe") — only one model per request is supported. For multi-model comparison, make one parallel tool call per model using the appropriate provider-specific tool.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
metno_forecastMET Norway ForecastARead-onlyIdempotent
Get weather forecast from Norwegian Meteorological Institute models. The models parameter is optional for this tool — omit it to use the default Met.no model. If specified, use canonical names such as "metno_nordic" or "metno_seamless". Only one model per request is supported. For multi-model comparison, make one parallel tool call per model using the appropriate provider-specific tool.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
seasonal_forecastSeasonal ForecastBRead-onlyIdempotent
Get long-range seasonal forecasts for temperature and precipitation up to ~7 months ahead.
Inputs: latitude (required), longitude (required), daily, hourly, +13 more.
weather_archiveWeather Archive (Historical)BRead-onlyIdempotent
Get historical weather data from ERA5 reanalysis (1940-present) for specific coordinates and date range.
Inputs: end_date (required), latitude (required), longitude (required), start_date (required), +12 more.
weather_forecastWeather ForecastARead-onlyIdempotent
Get weather forecast data for coordinates using Open-Meteo API. Supports hourly and daily data with various weather variables. When no models parameter is provided, the API automatically selects the best model for the given location (recommended). Only one model per request is supported — for multi-model comparison, use provider-specific tools with parallel calls.
Inputs: latitude (required), longitude (required), tilt, daily, +20 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 17 tools
The provider-specific forecast tools (gfs_forecast, ecmwf_forecast, jma_forecast, dwd_icon_forecast, gem_forecast, metno_forecast, meteofrance_forecast) overlap heavily with the generic weather_forecast, which also returns model data. Descriptions clarify that provider tools are for multi-model comparison, but the boundary between picking weather_forecast and a specific provider tool is not always obvious. Other tools (air_quality, marine_weather, elevation, flood_forecast, geocoding) are clearly distinct.
Names are consistently lowercase snake_case, which is good, but the pattern is mixed: some use {provider}_forecast (gfs_forecast), others use bare domain nouns (air_quality, elevation, geocoding), and weather_forecast/weather_archive use a weather_ prefix. Still readable and mostly predictable.
17 tools is on the heavy side, and seven of them are near-duplicate provider-specific forecast endpoints that largely could have been a single parameterized forecast tool. The remaining domain tools (air quality, marine, flood, elevation, geocoding, archive, seasonal, climate) each earn their place.
Coverage is broad and complete for a weather API: current/forecast, ensemble, seasonal, climate projections, historical archive, air quality, marine, flood, and elevation. Minor gaps like no explicit 'current conditions' or station/observation tool, but core lifecycle is well covered.
Maintenance
Related MCP Connectors
Global weather via Open-Meteo: forecast, historical, marine, air quality, geocoding, elevation.
US weather & geo for AI agents: forecasts, alerts, earthquakes, elevation, geocoding. No keys.
US weather & geo for AI agents: forecasts, alerts, earthquakes, elevation, geocoding. No keys.
Geocoding, routing, isochrones, traffic, weather, and place search for AI agents. 19 MCP tools.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables AI agents to access real-time and historical weather data through multiple weather APIs including OpenMeteo, Tomorrow.io, and OpenWeatherMap. Provides comprehensive meteorological information including current conditions, forecasts, historical data, and weather alerts.-
- AlicenseAqualityDmaintenanceProvides comprehensive access to Open-Meteo weather APIs, including forecasts, historical data, air quality, marine weather, and geocoding, enabling LLMs to retrieve weather information and location data.17MIT
- AlicenseAqualityFmaintenanceProvides comprehensive weather data (forecasts, historical, air quality, marine) from Open-Meteo through 12 tools, with support for single and batch location queries.121Apache 2.0
- AlicenseAqualityDmaintenanceProvides weather forecasts and geocoding lookup using free Open-Meteo APIs, enabling LLMs to query real-time weather and multi-day forecasts for any location.1ISC