---
format: "aidr-story-markdown/v1"
id: "51edf4c643816e6fc69aba606f22e49e530ea0b14e4cd171d8658ddbb9eb6a75"
canonical_url: "https://aidr.today/51edf4c6?lang=en"
title: "Introducing Timestral"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-15T12:00:00.000Z"
category: "Models"
topics: ["llm","open-source"]
source_urls: ["https://mistral.ai/news/timestral/","https://huggingface.co/mistralai/Timestral-123M","https://huggingface.co/mistralai/Timestral-28M"]
summary: "Timestral is a probabilistic forecasting model available in four sizes ranging from 5M to 209M parameters, designed to be CPU-efficient and zero-shot ready."
---

# Introducing Timestral

> [Open the canonical story](<https://aidr.today/51edf4c6?lang=en>)

**Published:** 2026-10-15T12:00:00.000Z
**Category:** Models
**Topics:** llm, open\-source

## Summary

Timestral is a probabilistic forecasting model available in four sizes ranging from 5M to 209M parameters, designed to be CPU\-efficient and zero\-shot ready\.

## Sources

- [Story source](<https://mistral.ai/news/timestral/>)
- [Story source](<https://huggingface.co/mistralai/Timestral-123M>)
- [Story source](<https://huggingface.co/mistralai/Timestral-28M>)