Laminar: pricing and features
by Laminar (Y Combinator S24)
Checked 2026-10-20 on Laminar's pricing page and GitHub repository. Anything we could not confirm says “not verified”. How we review.
Short answer: Laminar is an Apache-2.0 observability platform built for AI agents: OpenTelemetry-native tracing, plain-English “Signals” that flag agent behaviours, evals, SQL over traces and dashboards. Free tier: 1 GB of data with 7-day access. Paid plans: Starter $30/month (3 GB, then $2/GB) and Pro $150/month (10 GB, then $1.50/GB).
Plans
| Free | Starter | Pro | Enterprise | |
|---|---|---|---|---|
| Price | $0 | $30/month | $150/month | Custom |
| Data included | 1 GB, no overage | 3 GB, then $2/GB | 10 GB, then $1.50/GB | Custom |
| Data access | 7 days | 30 days | 6 months | Custom |
| Projects / seats | 1 / 1 | Unlimited / unlimited | Unlimited / unlimited | Unlimited / unlimited |
| Signals | Not included | Token-billed, $5 one-time credit | Token-billed, $5 one-time credit | Custom |
| Support | Community (Discord) | Slack | Dedicated | |
| Deployment | Cloud | Cloud | Cloud | Custom limits, on-premise |
Features
| Feature | Laminar | Source |
|---|---|---|
| Licence | Apache-2.0 | GitHub |
| Self-hosting | Docker Compose (light and full stacks) | README |
| Tracing | OpenTelemetry-native SDK with auto-instrumentation for Vercel AI SDK, Browser Use, Stagehand, LangChain, OpenAI, Anthropic, Gemini and more; OTLP ingestion | README |
| Signals | Plain-English behaviour detection across agent runs, with Slack alerts | README |
| Evals | SDK and CLI for running evals locally or in CI, with a UI to compare results | README |
| Datasets and labeling | Datasets and labeling queues | Pricing |
| Query and dashboards | SQL over traces, custom dashboards, MCP and CLI access | README |
| Browser session recording | Listed on the pricing page | Pricing |
| Compliance | SOC 2 Type II and HIPAA are listed on the pricing page; which plans include them is not verified | Pricing |
Who should pick Laminar
- Pick it if you run long, multi-step agents (including browser agents) and want traces, session recordings and behaviour alerts in one open-source tool.
- Pick it if you prefer a GB-based meter with unlimited seats on paid plans.
- Look elsewhere if prompt management or large human-review workflows are the priority: compare Langfuse, LangSmith and Braintrust.
- Leaving Helicone? See the Helicone migration guide.
FAQ
How much does Laminar cost?
Free is $0 with 1 GB of data, 7-day data access, 1 project and 1 seat. Starter is $30/month with 3 GB included, then $2/GB, 30-day access and unlimited seats. Pro is $150/month with 10 GB included, then $1.50/GB, and 6-month access. Enterprise is custom and includes on-premise. Checked on laminar.sh/pricing, 20 Oct 2026.
What are Laminar Signals and how are they billed?
Signals let you describe an agent behaviour in plain English (for example, an agent stuck in a loop) and have Laminar flag runs where it happens. Signals are not on the Free plan; on paid plans they are billed by the tokens Laminar's own analysis uses: $0.05 per million input tokens, $0.01 per million cached and $0.30 per million output, with a one-time $5 credit.
Is Laminar open source?
Yes. The lmnr repository is Apache-2.0 and self-hosts with Docker Compose; the README recommends the managed platform or the full Compose file for production.
Does Laminar support OpenTelemetry?
Yes. Its README describes an OpenTelemetry-native tracing SDK, and the pricing page lists OTLP trace ingestion.
Laminar or Langfuse?
Both are open source and self-hostable. Laminar is aimed at agent traces, with browser session recording and Signals; it meters GB of data. Langfuse meters units (traces, observations, scores) and has a broader prompt-management and annotation feature set. Compare on your own volume.