Laminar screenshot

Laminar

Open-source observability, debugging, and evaluation for AI agents

Pricing

Pricing not verified

  • Laminar automatically captures LLM calls, tool calls, sub-agents, costs, and tokens for a run. The data is presented in a transcript view so teams can inspect what an agent did while working through a task.
  • Laminar Signals analyze agent runs to surface failure modes that teams may not have defined in advance. Similar failures are clustered into patterns, providing a higher-level view of repeated behavior across runs.
Visit Laminar

What is Laminar?

Laminar is an open-source observability platform for AI agents. It captures agent activity to help development teams investigate failures and understand behavior in production.

The trace view records LLM calls, tool calls, sub-agents, costs, and tokens in a readable transcript. This gives teams a shared record of how an agent processed a task and used its tools.

Laminar Signals analyze runs for unanticipated failure modes and group similar failures into patterns. The platform also supports turning resolved error clusters into evaluation datasets for regression checks after a change.

The product is suited to teams that operate or build agentic systems and need debugging data alongside evaluations. Its CLI and MCP capabilities are intended to let coding agents query traces, evaluations, and signals during investigation.

Agent trace visibility

Laminar automatically captures LLM calls, tool calls, sub-agents, costs, and tokens for a run. The data is presented in a transcript view so teams can inspect what an agent did while working through a task.

  • Laminar automatically captures LLM calls, tool calls, sub-agents, costs, and tokens, and shows it in a readable transcript view.

Failure discovery and clustering

Laminar Signals analyze agent runs to surface failure modes that teams may not have defined in advance. Similar failures are clustered into patterns, providing a higher-level view of repeated behavior across runs.

  • Similar failures are clustered into distinct patterns to give you a high level overview.

Evaluation datasets from fixes

Laminar can turn resolved error clusters into evaluation datasets. Teams can run those evaluations after a change to check for regressions and compare agent behavior during iteration.

  • Run evals after a change to catch regressions and iterate with confidence.
Data & AnalyticsAgent observabilityLLM tracingAgent debuggingAI evaluationOpen source

User reviews

No reviews yet. Be the first to review this tool.

Log in to write a review.

← Back to AI Tools

Join the Aivolut newsletter

Occasional practical AI tips, app updates, and offers. Unsubscribe anytime.