--- url: "https://predictiontoolkit.com/tools/metar-ws/" publisher: "PredictionToolkit" dateModified: "2026-11-03" --- [← Weather research]() ![METAR.ws logo]() Data feed # METAR.ws: Station Observations and Weather-Model Streaming API METAR.ws is a WebSocket and REST data service for METAR/SPECI observations and derived station-level weather models. It delivers updates as supported source data arrives, with raw reports, decoded fields and timestamps for weather-contract research. [**3.5**/ 5Editorial rating]() [Editorial assessment]() [Open METAR.ws]() metar.ws **Available on:** API **Pricing:** [See plans and costs ↓]() [Screenshots]() [Explore alternatives]() **Availability: public** [Weather research]() [Data APIs]() +2 more [Streaming data]() [Historical data]() ## Screenshots and interface Product preview1 / 2 The provider's introduction to streaming weather data. ## What is METAR.ws? Keep observation time separate from the time the service received a report. Decoded temperature, wind, visibility and flight category accompany the raw text; station coordinates and elevation help align the observation site with a model grid. Grid-point distance and source age add freshness context. Model highs and lows are forecasts, not measured settlement results. Model output can be detected, partial or day-complete, and source publication and model-run timing bound availability. Match the contract's station, time zone, units and rounding convention before comparing a forecast extreme with the eventual observation. ### Who it is for - Building weather-market monitors as a developer. - Comparing forecasts with station observations. - Integrating station data with source timestamps. ## Feed coverage and updates - **Decoded observations** Receive temperature, wind, visibility and flight category together with the raw report. - **Separate timestamps** Distinguish observation time from platform receipt time. - **Station metadata** Use coordinates and elevation when aligning stations and model grids. - **Model signals** Receive derived highs and lows as model runs become available. - **Model lifecycle events** Handle detected, partial and day-complete output as distinct states. - **Freshness context** Inspect grid-point distance and source age. - **Historical API** Retrieve past METAR/SPECI data within the selected plan's limits. - **WebSocket subscriptions** Select stations and event types over an authenticated connection. ## Monitoring and data workflows ### Monitor a weather contract Match its settlement station and units, then subscribe to the relevant observation events. ### Compare model runs Store the model, run time, freshness and forecast extreme alongside the eventual observation. ### Audit report timing Compare observation and receipt times rather than treating every newly received report as a new measurement. ## Pricing and total cost EUR plans separate sandbox testing, a limited station set and broader production access. Observation and station entitlements differ by plan; model testing should not be treated as unrestricted observation access. ### Sandbox Free - 1 API key - 1 WebSocket connection - Model testing across stations - Historical API: 1 request/second ### Starter €49/month; 7-day trial - 10 stations - 2 WebSocket connections - 1 API key - Historical API: 10 requests/second ### Professional €199/month - All stations - 5 API keys - 10 WebSocket connections - Historical API: 100 requests/second ## Feed access and configuration 1. Check coverage for the required station and model. 2. Create an API key. 3. Connect to the documented WebSocket endpoint. 4. Subscribe to the station identifiers and event types you need. 5. Retain raw reports, units and both timestamps. 6. Handle partial model runs and reconnections explicitly. ## Useful links ### What to check - The reference station, time zone and rounding convention must match the contract rule. - Sandbox and paid plans have different observation and station entitlements. - Source publication and model-run timing limit when data becomes available. A closer look ## Ratings & reviews Editorial rating **3.5**/ 5 0 user reviews PredictionToolkit editorial ### Our take on METAR.ws METAR.ws makes report timing inspectable by separating observation from receipt and supplying decoded fields. That is useful for an integration, but contract research still needs station alignment and plan coverage checked against the actual settlement source; forecast extremes remain a separate kind of evidence. **Assessment criteria:** Task usefulness, documented scope, setup clarity, and access and cost transparency. **Assessment date:** 2026-10-05 **Documentation reviewed:** 2026-10-02 Documentation-based assessment. [How we rate]() ### User reviews 0 The main score switches to user ratings after 5 published reviews. **Be the first to share your experience** A specific example helps the next person choose the right tool. ## Related tools [Browse category]() ![Open-Meteo generated forecast response, chart and export formats]() ![Open-Meteo logo]()[Open-Meteo]() Data feed**4.0**Editorial An external weather API for comparing models and reconstructing forecasts available at a historical decision. - Access API Also in Weather Research and Data for Prediction Markets. ![Wethr product overview]() ![Wethr logo]()[Wethr]() Data feed**3.5**Editorial Weather-market research combining station observations, forecast-model comparisons, bias tools, alerts and tiered APIs. - Access Web · API Compare a weather-trading research dashboard. ![Weather Edge temperature-signal workflow and ensemble inputs]() [Weather Edge]() Data feed**3.0**Editorial Weather-model research and free daily temperature-market signals for Kalshi. - Access Web · Email Also in Weather Research and Data for Prediction Markets. ## Frequently asked questions about METAR.ws ### Does it include raw METAR text? Yes. Raw reports accompany the decoded fields. ### Do I need to parse GRIB files myself? The model product supplies derived station-level signals. ### Is historical data available? Yes. It is provided through an API with rate limits. ### Why retain both timestamps? They distinguish when the observation occurred from when the service received it. In this profile