Weather Forecasting
1 / 5
Physics-constrained AI models like EPT-2 learn conservation laws from data and avoid the physically impossible outputs that generic machine-learning forecasts produce.
2 / 5
June 2026 benchmarks show EPT-2 outperforming ECMWF HRES on RMSE for wind, temperature, and solar radiation across all 0–240 hour lead times.
3 / 5
EPT-2e’s 30-member ensemble beats the 50-member ECMWF ENS mean on both RMSE and CRPS at virtually every lead time, evaluated on over 10,000 global stations.
4 / 5
Native any-Δt forecasting and single-GPU inference make EPT-2 roughly 25% faster and four orders of magnitude cheaper to run than traditional NWP or competing AI models.
5 / 5
Energy traders can benchmark EPT-2 and 25+ models on their own region and variables by booking a demo with Jua.
New 2026 benchmarks show Jua's EPT-2 outperforms ECMWF HRES & ENS on RMSE and CRPS for wind, temperature & solar at every lead time. Explore now.