SHANGHAI – October 30, 2026 -- Envision's Tianji weather foundation model ranked first overall in the latest AI Weather Quest, organized by the European Centre for Medium-Range Weather Forecasts (ECMWF), outperforming nearly 20 top teams worldwide.

Tianji takes first place on every assessed metric for the 19-25 day window

Tianji led the variable-averaged ranking as well as the individual categories of near-surface temperature, mean sea level pressure and precipitation. All results apply to the 19-25 day forecast window.

The competition tests models through continuous, real-time forecasting and verification. Teams submit probabilistic forecasts weekly for weekly mean temperature, weekly mean sea level pressure and weekly accumulated precipitation. Organizers score the submissions against subsequent actual weather and update a public leaderboard. Temperature and precipitation are assessed over predominantly land-covered grid cells, while mean sea level pressure is assessed globally.

Earlier wins span global benchmarks and Chinese grid operators

Tianji previously achieved a top-tier placement in the WeatherBench 2 global evaluation, on par with Google's WeatherNext2 model. It also took first place overall in the deterministic forecast track of the AI Weather Forecast Model Demonstration Program run by China's National Meteorological Center.

In energy applications, Tianji won State Grid's AI load forecasting competition and China Southern Power Grid's business competition.

Envision describes Tianji as a physics-AI model built on petabyte-scale data

Robert Zheng, head of Envision's Tianji Weather Foundation Model, called subseasonal forecasting, covering two weeks to two months ahead, "one of the hardest challenges in meteorology." He said the model draws on petabyte-scale, high spatiotemporal resolution data and a "space-air-ground" three-dimensional data network. Envision calls Tianji the world's first physics-AI weather model.

Running on high-performance computing, Tianji covers time scales from minute-level nowcasting to subseasonal trend assessment. Envision says it delivers significantly improved accuracy for wind speed and irradiance in key regions.