Contact for pricing · Free trial🇺🇸🇩🇪
Process optimization for steel, chemicals and cement with explainable machine learning and setpoint advice.
Anomalib overview and key facts (October 2026)
Open source library from Intel's Open Edge Platform for visual anomaly detection, with OpenVINO export.
Free
What is Anomalib?
Anomalib is an open source Python library for visual anomaly detection. Machine learning engineers and quality teams use it to train, benchmark and deploy models that find defects in images and video without large sets of labeled defect examples.
The library collects state-of-the-art anomaly detection algorithms for benchmarking on public and private datasets, built on PyTorch Lightning, with a Python API and a command-line interface. Trained models can be exported to OpenVINO for accelerated inference on Intel hardware. The project is maintained in Intel's Open Edge Platform organization on GitHub, licensed under Apache 2.0, and had about 6,200 stars and a latest release (v2.6.2) on September 11, 2026 when we checked.
Anomalib AI features and use cases
- Trains defect detection on good partsModels such as PatchCore and PaDiM learn from images of good parts and flag anything that deviates. Quality teams can set up visual inspection when defect samples are rare or keep changing.
- Compares algorithms on your own dataThe library runs several algorithms on the same dataset and reports comparable metrics. Engineers choose the model that fits their parts and cycle time on evidence rather than on published benchmarks alone.
- Deploys models at the edgeExport to OpenVINO lets trained models run fast on Intel CPUs and accelerators next to the production line. Image data can stay on site and inspection keeps up with line speed.
What sets Anomalib apart
- Free under Apache 2.0 and usable in commercial projects.
- Maintained by Intel's Open Edge Platform organization with frequent releases and active development.
Where Anomalib is strongest
United States Available internationally
Anomalib is an open source project maintained by Intel's Open Edge Platform organization and used internationally without a focus market.
Questions to ask before you buy Anomalib
Anomalib is a library, not a finished inspection system: you need Python and machine learning skills, camera setup, labeled test data and your own operation and maintenance. There is no vendor support contract; plan who updates versions and validates models. Deployment is optimized for Intel hardware via OpenVINO.
Anomalib reviews
No reviews yet. Be the first to share how your team uses it.
Every review shows its verification level. We read each one before it goes live. How we check reviews
Questions about Anomalib
Is Anomalib free for commercial use?
Yes. Anomalib is released under the Apache 2.0 license, which allows commercial use.
Which algorithms does Anomalib include?
The documented models include PatchCore, PaDiM, STFPM and EfficientAd, among others.
Who maintains Anomalib?
The project is maintained in Intel's Open Edge Platform organization on GitHub, with contributions from the community.
Anomalib update historyno updates yet
No changes recorded for Anomalib yet.
Changes the vendor made and confirmed user reports, each checked against the vendor's own pages. We review every tool monthly.
Alternatives to Anomalib
Tools in the same category (Quality inspection) with the most overlapping markets.
From $100 per month🇺🇸
No-code frontline operations platform with AI Composer, AI Chat, AI agents and computer vision.
Research paper🇺🇸
Meta paper on SAM 3, a model that detects, segments and tracks objects in images and video from short noun phrases or example images.
Contact for pricing🇰🇷
Industrial AI from Korea: the Runway platform, equipment anomaly detection and engineering drawing analysis.