# TimesFM-3: A zero-shot foundation model for multivariate forecasting

DevFeed: [TimesFM-3: A zero-shot foundation model for multivariate forecasting](<https://devfeed.tech/articles/timesfm-3-a-zero-shot-foundation-model-for-multivariate-forecasting-6898.md>)

Original publisher: [Read original article](<https://research.google/blog/timesfm-3-a-zero-shot-foundation-model-for-multivariate-forecasting/>)

Published: 2026-08-31T17:19:40Z

Content type: article

Language: en

Sources: [The latest research from Google](<https://devfeed.tech/sources/the-latest-research-from-google.md>)

Topics: [foundation-models](<https://devfeed.tech/topics/foundation-models.md>), [Time Series](<https://devfeed.tech/topics/time-series.md>), [Google](<https://devfeed.tech/topics/google.md>), [datasets](<https://devfeed.tech/topics/datasets.md>), [Transformer architecture](<https://devfeed.tech/topics/transformer-architecture.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [generalization in machine learning](<https://devfeed.tech/topics/generalization-in-machine-learning.md>)

Tags: [data-management](<https://devfeed.tech/tags/data-management.md>), [features](<https://devfeed.tech/tags/features.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [foundation-models](<https://devfeed.tech/tags/foundation-models.md>), [google](<https://devfeed.tech/tags/google.md>), [machine-intelligence](<https://devfeed.tech/tags/machine-intelligence.md>), [model](<https://devfeed.tech/tags/model.md>), [product](<https://devfeed.tech/tags/product.md>), [time-series](<https://devfeed.tech/tags/time-series.md>), [transformer-architecture](<https://devfeed.tech/tags/transformer-architecture.md>), [zero-shot](<https://devfeed.tech/tags/zero-shot.md>)

## AI overview

Google Research introduces TimesFM-3, a 330-million-parameter time-series foundation model designed for accurate multivariate forecasting in a single forward pass. Pre-trained on more than one trillion real-world and synthetic time points, it jointly models coevolving series and external covariates in zero-shot settings without task-specific fine-tuning.

## Source excerpt

Data Management