#5454th of 75 in ML frameworks and libraries
pyod
Python library of 61 anomaly detectors with fit/predict API and agent skill
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 10.1k
- License
- BSD-2-Clause
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- Python
9,693 on 31 Mar 202610,052 on 20 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
PyOD is a Python library for outlier detection covering tabular, time series, graph, text, image and audio data through a scikit-style fit/decision_function/predict API. Version 3 adds ADEngine, which selects, compares and assesses detectors automatically, plus an od-expert skill for Claude Code and Codex and an MCP server exposing ten stateless tools.
Who it is for: Data scientists and ML engineers building anomaly detection pipelines
Strengths
- 61 detectors across six modalities behind one fit/predict interface
- Classic API unchanged in v3, so existing code keeps working
- Optional MCP server and Claude Code/Codex skill for agent-driven detection
- Benchmark-backed model selection using ADBench, TSB-AD, BOND and NLP-ADBench
Weaknesses
- MCP investigate and iterate tools are stateful and deferred, not yet available
- Full multi-turn agentic flow needs Claude Code or Codex via the skill
- Models persist via joblib or pickle, with the usual pickle trust caveats
- README excerpt gives no RAM, GPU or per-detector scaling figures
What it needs
- Needs numba, SUOD, mcp (optional extra)
- Models: tabular, time series, graph, text, image
Also in ML frameworks and libraries
See all 75| Rank | Project | Score |
|---|---|---|
| 1 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 2 | pytorchPython tensor library with GPU support and tape-based autograd for deep learning | 94 out of 100 |
| 3 | ultralyticsPython library and CLI for training and running YOLO vision models | 92 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and reference training scripts | 91 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 90 out of 100 |