Weights & Biases

Weights & Biases (W&B) is the leading MLOps platform for machine learning experiment tracking and model management, serving 200,000+ ML practitioners at OpenAI, Toyota, and Samsung. Automatically tracks experiments to save git commits, hyperparameters, model weights, and sample predictions for full reproducibility. Features developer-first UI prioritizing ease of use, best-in-class visualization, and rapid bottom-up adoption. Acquired by CoreWeave in 2025, making it the presumptive choice for teams on that GPU cloud. Provides auditable end-to-end ML workflows with comprehensive governance for enterprise teams.