Imagine predicting the path of a deadly cyclone seven days in advance—and getting it right. That’s exactly what DeepMind’s GenCast does.
(Here’s why this matters for all of us.)
GenCast, DeepMind’s new AI ensemble model, forecasts weather with state-of-the-art accuracy up to 15 days in advance. It even outperforms the European Centre for Medium-Range Weather Forecasts’ (ECMWF) system, which has been the gold standard.
In a head-to-head test of 1,320 weather variables, GenCast won 97.2% of the time. And when it comes to predicting extreme weather—like heat waves or hurricanes—it’s even more impressive.
The best part?
Unlike traditional models that require massive supercomputers, GenCast runs its entire ensemble forecast in just 8 minutes on a single TPU.
Faster. Smarter. More accessible.
DeepMind is releasing GenCast’s code and weights to the public, opening up new possibilities for researchers, governments, and even renewable energy planners.
In short: It’s not just about better weather forecasts. It’s about saving lives, cutting costs, and driving sustainable solutions.
Excited to see where this takes us? Repost this 
if you agree that smarter forecasts mean a safer world.