Weights & Biases is an AI developer platform for tracking machine learning experiments, building LLM applications, and evaluating and monitoring AI models in production.
About
Weights & Biases (W&B) is an AI developer platform that helps machine learning teams train models and ship AI applications with confidence. It addresses two common problems: losing track of experiments during model training, and lacking visibility into how LLM-based applications behave after they are built.
The platform combines experiment tracking for traditional model training with W&B Weave, a toolkit for building agentic AI applications. Weave records traces of LLM calls, retrieval steps, and agent actions, supports evaluations, provides a playground for exploring prompts and models, and offers monitors for improving applications in production. Integration is designed to be minimal: developers add a library import and initialize it with a project name, often a single line of code. The company states that thousands of organizations use W&B as their system of record for AI development, and published case studies include Canva, Microsoft, and Toyota. Weights & Biases is now part of CoreWeave Forge.
Key Features
Pricing & Fees
- Pricing: Self-hosted plans include a free Personal tier ($0/month, one user seat, personal projects only, corporate use not allowed) and an Advanced Enterprise tier with custom pricing. Cloud-hosted pricing has moved to CoreWeave Forge and is not fully detailed on the pages reviewed.
- Free academic access: Privately-hosted W&B is free for academic research, with unlimited projects and teams plus 200GB of free cloud storage.
- Self-hosting: The server runs locally with Docker and Python installed, on Linux, macOS, or Windows.
- Compliance: Certified under ISO/IEC 27001:2022, ISO/IEC 27017:2015, and ISO/IEC 27018:2019, and compliant with SOC 2 and HIPAA; the platform also supports NIST 800-53 and GDPR alignment.
- Enterprise features: Single sign-on, custom roles, audit logs, customer-managed encryption keys, and automated user provisioning on the Advanced Enterprise plan.
Who It's For
Weights & Biases is aimed at machine learning engineers, AI application developers, and research scientists who need to track training runs, evaluate LLM applications, or monitor AI systems in production. Typical use cases listed by the vendor include fine-tuning and training LLMs, retrieval-augmented generation (RAG), computer vision, AI governance, and quantitative trading. It serves industries such as autonomous vehicles, financial services, healthcare and life sciences, public sector, and scientific research. Academic researchers can use the privately hosted version at no cost.
FAQ
What is Weights & Biases?It is an AI developer platform for tracking machine learning experiments, building and evaluating LLM applications, and monitoring AI systems in production.
Is there a free plan?Yes. A free Personal self-hosted plan exists for one user on personal projects only, and privately hosted W&B is free for academic research with 200GB of cloud storage.
Which frameworks does W&B integrate with?It integrates with LangChain, LlamaIndex, PyTorch, Hugging Face Transformers, Lightning, TensorFlow, Keras, Scikit-learn, and XGBoost, typically starting with a single line of code.
What is the difference between self-hosted and cloud-hosted W&B?Self-hosting runs on the customer's own infrastructure with full control over data, while cloud-hosted is a fully managed service; cloud-hosted pricing is now handled through CoreWeave Forge.
Is Weights & Biases compliant with security standards?The platform is certified under ISO/IEC 27001:2022, 27017:2015, and 27018:2019, is compliant with SOC 2 and HIPAA, and supports NIST 800-53 and GDPR requirements.
Related Tutorials
View all →
Related Tools
View all →
OpenRouter OpenRouter is a unified, OpenAI-compatible API that gives developers pay-as-you-go access to 500+ AI models for text, image, video, and audio from 80+ providers. 
LangChain LangChain provides a comprehensive suite of products to support developers throughout their entire development workflow. 
LiteLLM LiteLLM is an open-source AI gateway and Python SDK that puts 100+ LLM providers, MCP servers, and agents behind one OpenAI-compatible API, with budgets, routing, and spend tracking for platform teams. 
Tokenhot Tokenhot is a unified LLM API gateway that puts 100+ text, image, video and TTS models from 30+ providers behind one OpenAI-compatible endpoint with pay-as-you-go billing. 
ZenMux ZenMux is an enterprise LLM gateway that unifies access to 100+ AI models through one API, with intelligent routing and automatic compensation for subpar outputs.