Google launches WeatherNext 3: hourly AI weather forecasts at 5-km resolution
ymatias · x · 2026-09-04
Google announced WeatherNext 3, its most advanced global AI weather model, alongside build access and the Google Earth AI planetary intelligence initiative.
Key points:
- The model learns from global geostationary satellite data, bypassing the typical six-hour data lag of traditional numerical weather prediction
- It generates a new forecast every hour at up to 5-kilometer resolution, always grounded in the latest satellite observations
- This matters most for fast-developing critical weather, with improvements in precipitation forecasts
Related event: Google Launches WeatherNext 3, Its Most Advanced Global Weather AI Model(19 posts)→
More from Models
- Opus 5.5 tops OpenRouter spending in 20 of 29 task categories — mmurph · 2026-10-21
- Meta team builds open-source 30B LLM from scratch in months, hails 'second era' — TimDarcet · 2026-10-21
- DFlash drafter for Muse Glimmer-30B: 5 layers, blockwise 16-token spec decoding — TimDarcet · 2026-10-21
- EvoPress quantization shrinks 30B model from 56GB to 19GB at near-bf16 accuracy — TimDarcet · 2026-10-21
- Radar chart: Muse Glimmer-30B benchmarked against Gemma4-31B and Qwen3.6-27B — TimDarcet · 2026-10-21
- Privacy evals: Muse Glimmer-30B leaks less than Qwen3.6-27B on CIMemories — TimDarcet · 2026-10-21