# Mistral AI releases Timestral forecasting model family, beats Google TimesFM-3

Digest AI · Generative AI & Models · published 2026-10-19T01:18:17Z

Canonical: https://digestai.news/story/mistral-ai-releases-timestral-forecasting-model-family-beats-google-ti

## Summary

Mistral AI announced Timestral, a family of four time‑series foundation models ranging from 5 M to 209 M parameters. The models are pretrained on a large synthetic corpus and can produce probabilistic forecasts zero‑shot. The 209 M checkpoint achieved the top rank among pretrained zero‑shot models on three public benchmarks – GIFT‑Eval, TIME and fev‑bench – according to Mistral’s own evaluation. It posted a fev‑bench SQL skill score of 50.39, beating Google’s TimesFM‑3 score of 48.99, and a GIFT‑Eval MASE of 0.658, narrowly ahead of t0‑gamma’s 0.661.

All four models run on standard server CPUs without a GPU; the 5 M model forecasts a day ahead in 10 ms, while the 209 M model needs 140 ms on an 8‑thread CPU. The models accept up to 32,768 time steps of history, output 21 quantiles per step, and support past‑only, known‑future and categorical covariates. Weights are released under a modified MIT licence and can be installed via pip from the public GitHub repository; fine‑tuning is available through Mistral Forge.

## Key points

- Timestral family includes 5M, 28M, 123M and 209M‑parameter models; 209M tops zero‑shot scores on GIFT‑Eval, TIME and fev‑bench.
- CPU‑only inference runs 10 ms (5M) to 140 ms (209M) for a 1‑day‑ahead forecast from three weeks of 15‑minute data.
- Model weights (837 MB fp32 for 209M) and code are open, installable via pip from Mistral’s GitHub repository.

## Why it matters

Provides high‑accuracy, zero‑shot time‑series forecasting on ordinary CPUs, lowering infrastructure costs and enabling developers to add scalable forecasts without GPU resources.

## Sources

1. [Mistral AI Releases Timestral: A 209M Forecasting Model Family That Outperforms Google’s TimesFM-3](https://marktechpost.com/2026/10/18/mistral-timestral-time-series-foundation-model) (MarkTechPost, 2026-10-19)

## Cite

Digest AI, "Mistral AI releases Timestral forecasting model family, beats Google TimesFM-3", 19 October 2026, https://digestai.news/story/mistral-ai-releases-timestral-forecasting-model-family-beats-google-ti

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/mistral-ai-releases-timestral-forecasting-model-family-beats-google-ti.json