198808xc/Pangu-Weather

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

1398 stars · 243 forks observed · 2026-10-06

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Official implementation of Pangu-Weather, a deep learning model for fast, accurate medium-range global weather forecasting using 3D neural networks, published in Nature. It provides pre-trained ONNX models and inference scripts for running forecasts on CPU or GPU.

Use cases

• run global weather forecasts with a pretrained AI model
• generate 10-day medium-range weather predictions
• experiment with 3D neural network weather models
• replace numerical weather prediction with deep learning inference
• reproduce Pangu-Weather research results

When to choose

• you need fast AI-based global weather forecasting without running a full NWP model
• you want to use the official pretrained Pangu-Weather ONNX models
• you are researching data-driven weather prediction

When to avoid

• you need local or regional high-resolution forecasting rather than global
• you need training code or full pipeline customization beyond released inference tools
• you need a maintained library with a license and active support

Facets

library · maturity maintenance

machine-learning deep-learning llm-inference artificial-intelligence machine-learning weather data-science python weather-forecasting onnx 3d-transformer research-code numerical-weather-prediction linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="198808xc/Pangu-Weather")