Pricing

Choose your path

Self-serve for individual developers, or enterprise deployment for regulated teams.

For developers

Developer

Free

  • —BYOK (bring your own LLM key)
  • —500 gateway requests per month
  • —Presidio + QI PII pipeline
Sign up free

Pro

Waitlist

  • —BYOK (bring your own LLM key)
  • —50,000 gateway requests per month
  • —Presidio + QI PII pipeline
  • —Audit log export
  • —Lucairn Certificates

All plans include BYOK (bring your own LLM key) and the Presidio + QI PII pipeline.

We measure gateway requests — a single call to Lucairn, regardless of upstream LLM token count. We do not meter LLM tokens; those are your direct spend with your BYOK provider.

How it works

1

Sign up

Create your free account and get a DSA API key in under a minute.

2

Add your LLM key

Bring your own Anthropic API key. Your key, your data, your control.

3

Start building

Paste the config into Claude Desktop, route the OpenAI SDK, or change the base_url in the Anthropic SDK.

MCP setup guide→Anthropic SDK setup guide→OpenAI SDK setup guide→SDK integration→
For enterprises

For enterprises

From Assessment to Production

Lucairn deploys through a proven three-step process. Start with an assessment of one workflow, prove value in a pilot, then scale to production on your infrastructure.

Step 1

Assessment

One workflow assessed. Exposure report with PII detection results. Architecture fit analysis. Pilot scope definition. 2–4 weeks.

Step 2

Pilot

One workflow in production. One entry pack deployed. Light customer tuning. Success criteria defined and measured. 8–12 weeks.

Step 3

Production

Customer-controlled deployment. BYOK default — your infrastructure, your encryption keys. Audit and evidence pipeline. Annual license.

Enterprise add-ons

Custom-trained PII shield

Fine-tune the level-3 PII shield on your de-identified domain corpus. Catches industry-specific identifiers a generic matcher misses — internal account formats, case classifications, free-text quasi-identifiers. The training runs in your environment on dedicated GPU; data and model weights stay with you. Priced per scope as part of the Enterprise engagement.

Every deployment starts with an assessment. Pricing depends on scope, workflows, and deployment model.

All core components are source-available. Inspect the code, evaluate the architecture, run it yourself. The managed service adds hosting, support, and operational expertise.

SDK source on GitHub: github.com/Declade/lucairn-sdks — gateway, sanitizer, witness, and VeilVault repositories are made available to evaluating customers under NDA during private preview; public-by-default rollout tracks the Stage 3 brand cutover.