Experiment tracking and model monitoring for ML teams
Weights & Biases, usually called W&B, is a platform for tracking machine-learning experiments. A few lines of code log metrics, hyperparameters, system usage and artifacts from a training run, and everything appears in a shared dashboard where runs can be compared and filtered. It also covers hyperparameter sweeps, dataset and model versioning, and reporting. The differentiator is reproducibility: teams use it as the record of what was actually tried, which is hard to reconstruct from terminal logs and notebooks after the fact.
Last updated: 2026-09-20. This site only provides an index; for exact features, pricing, and licensing, see the official website.
experiment tracking, hyperparameter search, model versioning, ml team reporting
If you're comparing similar products, check the alternatives below, or browse all tools in the AI Models & Platforms category.
Metrics over time, hyperparameters, code version, system resource use and output artifacts for each run. The value is being able to compare runs and reproduce a result weeks later.
Yes, W&B is free for personal projects and academic research, with limits on storage and collaboration features. Team and enterprise plans bill per seat and raise those limits.
It integrates with PyTorch, TensorFlow, Keras, scikit-learn and many others, and you can log manually from plain Python. Adding it to an existing training script is usually a few lines.
Developer console and API keys for the Claude model family
Node-based local interface for running image and video diffusion models
Browser playground for prompting and prototyping with Gemini models
Hosted Jupyter notebooks with optional free GPU and TPU runtimes