- Home
- AI Agent Observability Tools
- W&B Weave
W&B Weave
San Francisco, California, United States
What it is
W&B Weave gives teams tools to observe and improve production AI agents. It organizes traces into sessions and turns, with steps, tools, and sub-agents represented as distinct elements. Built-in and custom signals capture and classify agent interactions, and alerts can go to Slack or webhook automations. Its evaluation framework compares results and visualizes potential regressions before deployment; the Playground lets users compare models and prompting techniques against production traces. Pre-built scorers cover safety checks such as toxicity, bias, personally identifiable information, and hallucinations, alongside quality measures including coherence, fluency, and context relevance. Weave works with frameworks such as LangChain and LlamaIndex and models from OpenAI, Meta, Amazon Bedrock, and Anthropic. It offers Python and TypeScript libraries and accepts OpenTelemetry traces through an OTLP endpoint. The Free plan includes 1 GB of monthly data ingestion and 5 GB of storage. Pro is $60 a month, with 1.5 GB of monthly ingestion; extra Pro ingestion is $0.10 per MB. A 30-day trial is listed.
For and against
For
- Groups agent traces into sessions and turns.
- Routes alerts through Slack or webhook automations.
- Compares evaluations and visualizes possible regressions.
- Includes safety and quality scorers.
- Provides Python and TypeScript libraries.
Against
- Free plan limits ingestion to 1 GB monthly.
- Pro costs $60 a month.
- Extra Pro ingestion costs $0.10 per MB.
The verdict
W&B Weave combines agent tracing, evaluation, monitoring, and model comparison in one offering. Check the monthly ingestion limits and extra-usage rate when choosing between Free and Pro.
Who it suits
Weave suits teams building or operating production AI agents that need tracing, evaluations, interaction monitoring, or prompt comparisons. Its integrations and SDKs may fit teams using the listed LLM frameworks and models.
Plans and prices
Free
AI application evaluations · AI application tracing · AI application scorers · model experiment tracking
Read 20 Oct 2026 · site.wandb.ai
Pro
Unlimited teams · team access controls · service accounts · priority email and chat support
Read 20 Oct 2026 · site.wandb.ai
Enterprise
HIPAA-compliant option · private connectivity · customer-managed encryption keys · SSO
Read 20 Oct 2026 · site.wandb.ai
On the record
- Acquisition news
- CoreWeave announced on 4 March 2025 that it would acquire Weights & Biases and completed the acquisition on 5 May 2025.investors.coreweave.com · read 20 Oct 2026
- Agent-native tracing
- Traces are organized around sessions, turns, steps, tools and sub-agents as first-class objects.site.wandb.ai · read 20 Oct 2026
- Apple app requirements
- The W&B iPhone app requires iOS 18 or later, and its Mac version requires macOS 15 or later on Apple silicon.apps.apple.com · read 20 Oct 2026
- Behavior monitoring
- Built-in and custom signals classify agent interactions, while Slack notifications and webhooks route detected issues.site.wandb.ai · read 20 Oct 2026
- Company
- Weights & Biases says it was founded in 2017 by Lukas Biewald, Chris Van Pelt, and Shawn Lewis and was acquired by CoreWeave in May 2025.site.wandb.ai · read 30 Sep 2026
- Company facts
- LinkedIn lists Weights & Biases as privately held, founded in 2017, headquartered in San Francisco and employing 201–500 people.linkedin.com · read 20 Oct 2026
- Deployment options
- The deployment guide describes multi-tenant SaaS and single-tenant AWS Dedicated Cloud deployments for Weave.site.wandb.ai · read 20 Oct 2026
- Documentation depth
- The support center lists a dedicated Weave section with 16 articles and eight tags.docs.wandb.ai · read 20 Oct 2026
- Encryption
- W&B states that data is encrypted in transit using TLS 1.2+ and at rest using AES 256.site.wandb.ai · read 30 Sep 2026
- Enterprise controls
- Enterprise pricing lists options including single sign-on, audit logs, secure private connectivity, and customer-managed encryption keys.site.wandb.ai · read 30 Sep 2026
- Evaluations
- An imperative evaluation API compares application or model changes with visualizations intended to expose regressions.site.wandb.ai · read 20 Oct 2026
- Guardrails
- Prebuilt safety scorers cover toxicity, bias, PII and hallucinations, while quality scorers cover coherence, fluency and context relevance.site.wandb.ai · read 20 Oct 2026
- Integrations
- The deployment guide lists integrations for OpenAI, Anthropic, Cohere, Mistral, LangChain, LlamaIndex, DSPy, Google Gemini, Amazon Bedrock, Together and Groq.site.wandb.ai · read 20 Oct 2026
- Leaderboards
- Evaluation results can be aggregated into leaderboards and shared across an organization.site.wandb.ai · read 20 Oct 2026
- Legal name and ownership
- Crunchbase identifies the legal entity as Weights & Biases, Inc., founded by Chris Van Pelt, Lukas Biewald and Shawn Lewis, and records seven funding rounds and acquisition by CoreWeave.crunchbase.com · read 20 Oct 2026
- Licence
- The public Weave repository is released under the Apache-2.0 license.github.com · read 20 Oct 2026
- Limit: Apple app access
- An App Store reviewer complained that meaningful functionality is gated behind a paid multi-tenant account.apps.apple.com · read 20 Oct 2026
- Limit: G2 complexity and cost
- G2 reviewers repeatedly mention an overwhelming interface, documentation or guidance gaps, sluggish charts on large projects and rapidly rising storage or per-user costs.g2.com · read 20 Oct 2026
