APIbenchmarks
Open-Meteo logo

Open-Meteo

Open-Meteo · Ranked #4 of 7 in Weather APIs

79.3/ 100
BStrong

Open-source, key-free weather API beloved by developers; free for non-commercial use with no signup, paid reserved instances for production.

Best for

Free open-source weather API

Screenshot of Open-Meteo

Overview

Open-Meteo is a free, open-source weather API that aggregates numerical weather prediction data from 30+ models run by 15+ national weather services (ECMWF, NOAA GFS, DWD ICON, Meteo-France, JMA, UK Met Office and others) and exposes it through simple HTTP GET requests returning JSON, CSV or XLSX. It was created by Patrick Zippenfenig and is unusual in the weather-API market for being genuinely open: the full server source is on GitHub (~5.7k stars), all data sources are credited, and the raw weather datasets are mirrored on the AWS Open Data registry. The non-commercial tier requires no account or API key, which is the single feature developers cite most when comparing it to OpenWeatherMap or Tomorrow.io. It selects the best high-resolution local model for any given coordinate automatically, blending down to 1-2 km resolution in Central Europe, France, the UK and North America while falling back to 9-11 km global models elsewhere.

The product is aimed at indie developers, hobbyist apps, researchers and data scientists who want clean access to multi-model forecast and reanalysis data (1940-present) without wrangling GRIB/NetCDF files. Its breadth is a real differentiator: beyond standard surface forecasts it ships marine/wave, air-quality-with-pollen, solar radiation (including global tilted irradiance for PV panels), ensemble probabilistic forecasts, seasonal forecasts to 9 months, flood, climate-projection and historical-forecast endpoints, all under one consistent query grammar. Performance is strong for a free service, the vendor cites sub-10ms single-location forecast responses and sub-100ms multi-decade historical queries, helped by a FlatBuffers-based response encoding and Swift backend.

Where it loses: it is model data only, not station observations, so users wanting validated actuals or hyperlocal nowcasting (e.g. radar-based rain-in-the-next-minutes) must look elsewhere, and there is no formal SLA on the free tier. Commercial use legally requires a paid subscription ($29/$99/mo and up), and the free tier's rate limits (600/min, 5k/hour, 10k/day) draw occasional frustration from developers who hit them during bursty workloads. Paid plans add a dedicated endpoint, API key and a 99.9% uptime target, but the absence of a public third-party-audited status/SLA page and the small (effectively single-maintainer) team are the main reliability caveats for production buyers.

How this score is derived

The APIbenchmarks Index is a weighted sum of four dimensions, each scored on an absolute 0–100 reference scale. See the methodology for every mapping.

DimensionScoreWeightContribution
Documentation & DXInteractive docs with a live URL/chart builder, copy-paste code in Python/TS/Swift, and per-endpoint parameter references make it among the easier weather APIs to onboard.
82
30%24.6
ReliabilityPaid tiers advertise a 99.9% uptime target on reserved instances, but the free tier has no guarantee and there is no prominent public third-party status page or audited SLA history.
66
25%16.5
Ecosystem & SDKsOfficial SDKs in Python, TypeScript, Swift, Kotlin and Java plus community R/Dart packages, AWS Open Data mirrors, and an open-source server (~5.7k GitHub stars) give it a healthy footprint for a small project.
76
25%19.0
AccessibilityNon-commercial use needs no account, no API key and no auth, a single HTTP GET returns JSON, which is the lowest barrier to entry of any major weather API.
96
20%19.2
APIbenchmarks Index (ABI)79.3

Table 1. Derivation of the ABI for Open-Meteo. Contribution = score × weight; the index is their sum.

At a glance

Vendor
Open-Meteo
Pricing model
Free non-commercial / per-call commercial
Free tier
10k calls/day, no key (non-commercial)
Official SDKs
7 languages

Pricing

Free / Open-Access (non-commercial)$0No API key or account; rate-limited to 600 calls/min, 5,000/hour, 10,000/day (~300k/month). No commercial use, no uptime guarantee.
API Standard$29/mo1M API calls/month, API key + dedicated endpoint, commercial licence, 99.9% uptime target. Includes Forecast, Marine, Air Quality, Geocoding, Elevation, Flood APIs.
API Professional$99/mo5M API calls/month; adds Historical Weather, Ensemble Forecast and Climate Change APIs on top of Standard.
API EnterpriseCustom (contact info@open-meteo.com)50M+ calls/month or custom volume/enterprise contracts; price not published on the pricing page.

Key features

  • 30+ NWP models from 15+ national weather services (ECMWF, GFS, ICON, Meteo-France, JMA, UK Met Office, etc.)
  • Up to 1-2 km resolution in Central Europe/France/UK/North America; 9-11 km global
  • 16-day forecasts plus 80+ years (1940-present) of historical reanalysis
  • Marine/wave, air quality + pollen, and solar radiation (incl. global tilted irradiance for PV)
  • Ensemble probabilistic forecasts and seasonal forecasts to 9 months
  • Flood, climate-projection, geocoding and elevation endpoints
  • JSON, CSV and XLSX output with FlatBuffers-encoded responses for typed SDKs
  • Multi-location queries via comma-separated coordinates in one request
  • 15-minutely resolution data for Central Europe and North America
  • Open data: full server source on GitHub and datasets mirrored on AWS Open Data

Official SDKs

PythonTypeScriptSwiftKotlinJavaR (community)Dart (community)

Strengths & trade-offs

Strengths
  • +No API key or account required for non-commercial use, a single HTTP GET returns JSON
  • +Blends 30+ models from 15+ national weather services and auto-selects the best local high-res model per coordinate
  • +Exceptionally broad data: forecast, historical (1940-present), marine, air quality, solar/PV, ensemble, seasonal, flood, climate
  • +Fully open source server with all data sources credited; datasets also mirrored on AWS Open Data
  • +Flat, predictable commercial pricing ($29/$99) with no per-call overages or surprise charges
  • +Fast responses, vendor cites sub-10ms single-location forecasts and sub-100ms multi-decade historical queries
Trade-offs
  • Model data only, no real station observations, so no validated actuals or hyperlocal radar nowcasting
  • Free tier rate limits (600/min, 5k/hr, 10k/day) frustrate some developers with bursty workloads
  • Commercial use legally requires a paid plan even at low volumes
  • No public third-party-audited status page or detailed SLA history; only a 99.9% uptime 'target'
  • Effectively maintained by a very small team / single core maintainer, bus-factor risk for production use
  • Precipitation probability and some derived fields historically reported as missing or limited by users

What developers say

Developers widely praise Open-Meteo as accurate, generous and refreshingly key-free, with the main complaints being aggressive free-tier rate limiting and that it serves model data rather than real observations.

OpenMeteo is pretty amazing too, and doesn't require an account or API key, which is nice.

Key figures

Single-location forecast response timeunder 10 ms (typical)Open-Meteo Features page (vendor)
Multi-decade historical query response timeunder 100 msOpen-Meteo Features page (vendor)
Paid-tier uptime target99.9%Open-Meteo Pricing page (vendor)
Free-tier rate limit600 calls/min, 5,000/hour, 10,000/dayOpen-Meteo Pricing page (vendor)
Standard plan price / quota$29/mo for 1M callsOpen-Meteo pricing announcement (vendor Substack)
GitHub stars (open-meteo/open-meteo)~5,745 stars, 375 forksGitHub API

Compare Open-Meteo head to head

Sources

  1. https://open-meteo.com/
  2. https://open-meteo.com/en/pricing
  3. https://open-meteo.com/en/features
  4. https://open-meteo.com/en/docs
  5. https://openmeteo.substack.com/p/api-subscriptions-for-commercial
  6. https://github.com/open-meteo/open-meteo
  7. https://registry.opendata.aws/open-meteo/
  8. https://news.ycombinator.com/item?id=28499910
  9. https://news.ycombinator.com/item?id=28506606

Figures last verified 2026-06-27. Spotted an error? corrections@apibenchmarks.com