Google Releases TimesFM, a Time Series Foundation Model
mdancho84 · x · 2026-07-11
Google has released a time series foundation model, TimesFM.
It targets forecasting scenarios like sales trends, market prices, user traffic, energy demand, and cryptocurrency volatility, with a focus on zero-shot forecasting: making predictions without retraining on specific data. The post also mentions it was pre-trained on 100B real-world data points, can run locally, is free to use, and is licensed under Apache.
Related event: Google Open-Sources TimesFM for Zero-Shot Forecasting(2 posts)→
More from Research
- MCP fault injection study: agent task scores drop 60% under realistic tool failures, scale doesn't help — dair_ai · 2026-10-26
- Study: AI reviewing AI misses most reward hacking; best models miss 25%+ of hacks — ronbodkin · 2026-10-26
- FLUX.3 Klein step-distilled to 4-8 steps with CFG baked in — Fit-Structure-6489 · 2026-10-26
- NLI alone rejects 5/6 new agent memories; a 0.45 similarity gate fixes it all — Alternative_Pin9598 · 2026-10-26
- AI folks, stop saying "biology" when you mean drug development — KevinKaichuang · 2026-10-26
- ostris pre-training "Primal VAE": 1-channel latents match modern 8x VAEs via patching — ostrisai · 2026-10-26