#7575th of 126 in ML frameworks and libraries
anomalib
Library for training and benchmarking visual anomaly detection models
- Stars
- 6.3k
- License
- Apache-2.0
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- Python
5,539 on 4 Apr 20266,349 on 31 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Anomalib collects image and video anomaly detection algorithms such as Patchcore, Padim and Stfpm, with benchmark datasets like MVTec AD. It offers a Python API and an `anomalib` CLI for training, prediction and benchmarking, and most models export to OpenVINO. It also includes Anomalib Studio, a pre-release low-code web app for training and deploying models from cameras or image folders.
Who it is for: ML engineers building visual defect detection, e.g. industrial inspection
Strengths
- Many ready-made anomaly detection models and benchmark datasets behind one API and CLI
- Most models export to OpenVINO for inference on Intel hardware
- Install extras for CPU, CUDA 12.6/13.0, ROCm and Intel XPU
- Built-in benchmark command and Lightning loggers (W&B, Comet, TensorBoard)
Weaknesses
- Focused on visual anomaly detection; not for tabular or time-series data
- Anomalib Studio is pre-release and may be incomplete or unstable
- Intel GPU training supports a single GPU only
- v2.7.0 raised minimum PyTorch to 2.10+ and removed the legacy ONNX exporter
What it needs
- GPU optional
- Needs PyTorch 2.10+, Lightning 2.2+, OpenVINO 2024.0+ (optional)
- Models: Patchcore, Padim, Stfpm, WinCLIP, MH-PatchCore
- port 3000
Quick answers
Is anomalib open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Does anomalib need a GPU?
No. It runs on a CPU and can use a GPU when you have one.
Also in ML frameworks and libraries
See all 126| Rank | Project | Score |
|---|---|---|
| 1 | pytorchPython tensor library with GPU support and tape-based autograd for deep learning | 95 out of 100 |
| 2 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 3 | ultralyticsPython library and CLI for training and running YOLO vision models | 93 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and reference training scripts | 93 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 91 out of 100 |