Google releases TimesFM-3, a 330M multivariate zero-shot forecasting foundation model
AxSaucedo · x · 2026-09-10
Google Research releases TimesFM-3 (330M parameters, pretrained on 1T+ time points), its first time-series foundation model with native multivariate forecasting. Unlike TimesFM-2.5 and earlier (univariate only), it jointly predicts coevolving series with covariates in a single forward pass, claims state-of-the-art zero-shot results on major benchmarks, and is open-sourced on GitHub and Hugging Face.
More from Models
- vLLM Ships Full Support for DeepSeek-V4.1-Flash's New Architecture — vllm_project · 2026-09-10
- DeepSeek-V4.1-Flash reportedly offers continuous reasoning effort from 1 to 100 — zainhas · 2026-09-10
- DeepSeek multimodal team praised for unusually careful pre-training data work — zephyr_z9 · 2026-09-10
- DeepSeek v4.1 flash evals leak: SOTA on deepswe, competitive on terminal bench — zainhas · 2026-09-10
- A new kind of encoder-decoder model has appeared, researcher says — bclavie · 2026-09-10
- DeepSeek V4.1 Flash has just 20 decoder layers, shallower than its predecessor — teortaxesTex · 2026-09-10