Google AI Releases TimesFM-3: A 330M Parameter Zero-Shot Foundation Model For Multivariate Time Series Forecasting

Google Research released TimesFM-3, a 330 million parameter time series foundation model pretrained for multivariate forecasting on over 1 trillion time points. Unlike TimesFM 2.5 and earlier univariate checkpoints, it jointly forecasts multiple targets and accepts past and past-future covariates zero-shot. It ranks first among foundation models on GIFT-Eval, fev-bench, and TIME. Code is Apache-2.0, but weights carry a non-commercial, non-production license; TimesFM 2.5 remains the Apache-2.0 option.