Skip to main content
Laminar is an open-source observability platform for AI agents that proactively analyzes every trace and alerts you when a run goes wrong. Our fine-tuned model flow-1 delivers frontier-level trace analysis at a cost that makes it practical to monitor every run without sampling. These findings help you turn real failures into eval cases, discover unknown failure modes, and prioritize improvements based on where users struggle. Get started with Laminar Cloud or self-host Laminar on your own infrastructure.
Traces stream from your agent in production into flow-1, which reads every one of them; the count of traces read goes up with each. Most traces have no issue. Failing ones turn orange and land in Issues: Wrong refund amount grows from 3 to 5, Report contradicts tests from 1 to 2, and a new issue, Stuck retrying login, is discovered. The top issue then goes round the improvement loop: its traces are turned into eval cases, the fix is shipped to the agent, and the issue is marked fixed.

flow-1 reads every trace your agent produces. Failures collect into issues, including ones you didn't know about, and the top issue goes back to the agent through eval cases and a fix.

Tracing

Set up comprehensive tracing for your agents in minutes. First-class support for AI SDK, Claude Agent SDK, LangChain, Browser Use, and much more.

Signals

Set up Signals to automatically catch errors in your agent runs. Describe what to flag in plain language, spot recurring issues and get notified in Slack or email.

Debugger

Record a run, replay it from cache up to the point you’re testing, and prove the fix with an eval. Built for a coding agent to run.

Evaluations

Run evals against datasets locally or in CI. Catch regressions before they ship.

Datasets & Labeling

Build evaluation datasets from production traces.

Playground

Replay any traced span. Swap prompts or models and compare results side by side.

Platform

Realtime traces, full-text search, raw SQL access from the UI, CLI, or MCP, and custom dashboards.