# Google previews TimesFM 3.0 for forecasting related data series in BigQuery

_Published Monday, October 12, 2026 at 12:42 PM EDT · AI · Latest · Tier 2 — Notable_

![Google previews TimesFM 3.0 for forecasting related data series in BigQuery — Primary](https://storage.googleapis.com/gweb-cloudblog-publish/images/09_-_Data_Analytics_tFH57V6.max-2600x2600.jpg)

Google announced a preview of TimesFM 3.0 in BigQuery, adding forecasts that jointly predict related data series and use external drivers such as holidays and promotions. Earlier versions forecast from the history of a single series alone. The 330-million parameter model runs through BigQuery's AI.FORECAST SQL function without requiring users to train custom models.

Users can supply historical signals and drivers known across the future forecast period. BigQuery Conversational Analytics can also generate and execute these forecasts from natural-language requests. TimesFM 3.0 supports existing single-series forecasting and anomaly detection workflows, while the new multivariate capability is in preview.

## Sources

- [Cloud Blog](https://cloud.google.com/blog/products/data-analytics/bigquery-timesfm-3-0-with-multivariate-time-series-forecasting/)

---
Canonical: https://techandbusiness.org/newswire/PmPY1clzTYKCXZH1ZIGNsS
Published: 2026-10-12T16:42:21.140Z
Story chronology: 2026-10-12T16:00:00.000Z
Retrieved: 2026-10-12T19:43:08.749Z
Publisher: Tech & Business (techandbusiness.org)