MMLForge Search models, datasets, tasks… Start free
Home › Models › Time Series Forecasting › Datadog/Toto-2.0-4m
T

Datadog/Toto-2.0-4m

Time Series Forecasting·Datadog· 13.0K· 18
pytorch apache-2.0 Time Series Forecasting 4.1M params arxiv:2602.12147arxiv:2605.20119

Toto (Time Series Optimized Transformer for Observability) is a family of time series foundation models for multivariate forecasting developed by Datadog. Toto 2.0 is the current generation, featuring u-μP-scaled transformers ranging from 4m to 2.5B parameters, all trained from a single recipe. Forecast quality improves reliably with parameter count across the family.

Open in MLForge Sign up free Desktop app Source ↗
# pull & run locally
pip install mlforge-sdk && mlforge pull Datadog/Toto-2.0-4m

Model details

Task
Time Series Forecasting
Provider
Datadog
Framework
pytorch
Parameters
4.1M
Size
17 MB
License
apache-2.0
Downloads
13.0K
Likes
18
Paper
arXiv:2602.12147
Updated
2026-06-04

About Datadog/Toto-2.0-4m

Toto (Time Series Optimized Transformer for Observability) is a family of time series foundation models for multivariate forecasting developed by Datadog. Toto 2.0 is the current generation, featuring u-μP-scaled transformers ranging from 4m to 2.5B parameters, all trained from a single recipe. Forecast quality improves reliably with parameter count across the family.

Related Time Series Forecasting

C amazon/chronos-2 Time Series Forecasting ·119.5M params 15.5M 340 🤗 HF C autogluon/chronos-bolt-small Time Series Forecasting ·47.7M params 13.7M 44 🤗 HF C autogluon/chronos-2 Time Series Forecasting ·119.5M params 8.2M 33 🤗 HF C autogluon/chronos-2-small Time Series Forecasting ·27.9M params 2.9M 4 🤗 HF C autogluon/chronos-bolt-tiny Time Series Forecasting 2.5M 13 🤗 HF K NeoQuasar/Kronos-Tokenizer-base Time Series Forecasting ·4.0M params 2.5M 60 🤗 HF

Browse all Time Series Forecasting models →