Introducing Timestral

Mistral AI and 1 other source

NewsPrimary source

Published directly by Mistral AI and picked up by 1 other source.

Mistral AI released Timestral, a family of probabilistic forecasting models available in four sizes ranging from 5 million to 209 million parameters. The models are designed to be CPU-efficient and work without fine-tuning.

Why it matters

Probabilistic forecasting models that run efficiently on standard hardware could enable time-series prediction tasks across more organizations and applications.

Key points

Four model sizes available (5M, 7M, 8M, and 209M parameters)
CPU-efficient operation
Zero-shot capability out of the box

Original source

Mistral AIFirst-party· Oct 15, 12:00 PM

Introducing Timestral

Then picked up by

Oct 19, 09:59 AM Hacker News · discussionTimestral: Mistral release time-series foundation models

Why is this ranked here?

Age	-26.8
Community engagement	9.6
Content type	2
Corroborating sources	4
Primary source	20
Total	8.8

These components add up to 8.799999999999997, not 8.8. The total is what ordered your brief; the difference means a component was not recorded, so treat this breakdown as incomplete rather than as the whole reason.