GribStream
Live & Historical Weather Forecast API
Query the latest forecasts from NOAA, ECMWF, CWA, and AI systems as they become available, then use the same API to backtest against years of forecast history. GribStream lets your application ask for exactly the weather data it needs: coordinates, time ranges, calculated fields, condition filters, and CSV, JSON, or NDJSON output.
Built for fast, predictable forecast access
Three pillars that keep the data fast, the API simple, and the pricing predictable.
Fast at production scale
Fetch fresh forecasts or months of hourly history for tens of thousands of points in one HTTP request, in seconds.
API-first workflow
Simple API, open-source clients and clear documentation. Try it in minutes.
Predictable usage
Predictable pricing and high throughput. Query only the forecast data you need, when you need it.
Developer workflow
One request, exactly the forecast data you need
Pick a dataset, coordinates, variables, and time range. GribStream returns tidy rows for applications, ML pipelines, dashboards, and research notebooks.
{
"fromTime": "2025-05-01T00:00:00Z",
"untilTime": "2025-05-02T00:00:00Z",
"coordinates": [{ "lat": 40.758, "lon": -73.985 }],
"variables": [{ "name": "TMP", "level": "2 m above ground", "info": "" }]
}
Platform capabilities
Query current and historical forecasts with advanced filtering, expressions, and ensembles.
Latest forecasts and history
Query current forecasts, then use the same API for multi-year forecast history and backtesting.
Flexible formats
Choose CSV, JSON, or NDJSON for streaming workflows.
Precise locations
Query any coordinates with flexible precision.
Time controls
Request ranges or explicit timestamps.
Best-run selection
Automatic selection of the best eligible run.
asOf model-run-time cutoff?
Restrict queries to model runs generated at or before a chosen model-run timestamp.
Derived fields
Compute derived fields on the fly.
Condition search
Filter by conditions instead of raw grids.
Ensemble support
Full ensemble members and statistics.
Expert support
Prompt, precise support from engineers who know the data.
Forecast coverage
Operational, regional, ensemble, analysis, and AI forecast systems exposed through one consistent API. Use this page for a quick scan, then open the full catalog for dataset-level details.
Category
AI & Research
Active AI forecast systems for deterministic, ensemble, and rapid scenario workflows.
Provider
NOAA
Operational, analysis, ensemble, and seasonal families across regional and global domains.
Provider
ECMWF
Operational, ensemble, wave, and AI forecast families sourced from ECMWF Open Data.
Provider
Taiwan CWA
Regional WRF guidance for East Asia and the western Pacific.
Pricing plans
A Free Plan is available to get you started. Upgrade to one of our Pro plans for higher daily usage limits.
For larger-scale usage, customized plans, heavy backfills, higher performance, or additional features, contact us.
Pro Plan
Daily Limit: 48,000 credits
$9.90 / month
Pro Plan 2x
Daily Limit: 96,000 credits
$18.80 / month
Pro Plan 4x
Daily Limit: 192,000 credits
$35.70 / month
Pro Plan 8x
Daily Limit: 384,000 credits
$67.80 / month
Pro Plan 16x
Daily Limit: 768,000 credits
$128.80 / month
Pro Plan 32x
Daily Limit: 1,536,000 credits
$244.70 / month
Per request quota calculation
Credits are calculated from what is actually returned.
Credits = returned_valid_times
* parameters
* ceil(coordinates / 500)
In words: valid times returned, times the number of weather parameters, times bundles of 500 coordinates.
We do not charge for times that are not returned, for example model-specific skipped horizons.
Examples
- If your response returns 24 valid times, with 2 parameters, for 700 coordinates: 24 * 2 * 2 = 96 credits
- If your response returns 100 valid times, with 1 parameter, for 3,500 coordinates: 100 * 1 * 7 = 700 credits
Cache hits are billed at 10 % of normal credits.
Ready to build
Start with a free API token
Create an account, run the Quick Start, or browse forecast coverage before wiring GribStream into your application.
