Project AURIONAI Use-Cases Research & Innovation
IBM
time series

Granite TSPulse

IBM tiny time-series model for anomaly detection and more (paper 19 May 2025, ICLR 2026). Claims #1 on TSB-AD univariate and multivariate (+20%) with no scores on the card. Maximo integration unverified.

IBMIBMUSA
Architecture overview of TSPulse from the model card.
Levelmodel
Classtime series
Parameters1.08M
Contextnot in the notes
Released19 May 2025
Opennessopen weights permissive
Runs onCPU server
Statusnot in the notes

Benchmark Results

vendor-reported
TSB-ADtime series
IBM claims #1 on univariate and multivariate, +20%; no scores published on card; conflicts with PaAno rank 1vendor-reported
no score in the notes
rank 1huggingface.co
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).

Industry Use Cases

Industrial
Anomaly detection on sensor, SCADA and historian data

R&D Topics

Anomaly detection and predictive maintenance on sensor dataexemplarTime-series foundation modelsproduced by

Identity

Developer
IBM
Family
not in the notes
Released
19 May 2025 sourcesnotes
Weights published
–
Record checked
10 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
not in the notes Dual time and frequency-space pretraining
Total parameters
1.08M sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
time series
Output
time series
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights permissive
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
CPU server
Footprint
CPU, GPU-free
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Lineage

The notes record no lineage for this model.

Sources

  1. ibm-granite/granite-timeseries-tspulse-r1 , Hugging Face huggingface.co https://huggingface.co/ibm-granite/granite-timeseries-tspulse-r1
  2. IBM's Tiny TSPulse Beats Models 100x Bigger Across 75 Datasets | AlphaSignal alphasignal.ai https://alphasignal.ai/news/ibm-s-tiny-tspulse-beats-models-100x-bigger-across-75-datasets

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Granite TSPulse", https://aurion.semai.id/models/ibm-granite-tspulse/, records of October 2026.