Systems nominal 7.5M executions / 30d
Pricing · per-call, no seats

Pay for calls. Not for seats. Not for CPUs.

Every plan includes the hosted runtime, dashboard, Playground, and every engineer on your team. We quote per deployment — call volume, regions, self-host — in one scoping call.

Free pilot

Evaluate with your own workbook.

Not a demo dataset — your rating or pricing workbook, compiled on the managed runtime and validated against results you already trust.

  • Hosted runtime · managed SaaS
  • Automatic I/O discovery and validation
  • Regression tests from your approved quotes
  • Playground + MCP server for agents
  • Direct line to the engineering team
Request pilot access
Production

Enterprise

Production, on your terms.

A flat annual agreement sized to your call volume, deployment model, and support level — quoted after one scoping call.

  • Unlimited models · prod, staging, dev
  • Private execution destination in your AWS
  • OIDC SSO + SCIM 2.0
  • Custom audit retention
  • Response SLA · Slack Connect support
  • DPA + penetration-test reports under NDA
Talk to sales

Self-serve plans open later — production starts sales-assisted.

How pricing works

Three rules. No fine print.

01

Per call

A call is one successful execution against a compiled model. Validation failures, rejected requests, and health checks are not counted.

02

No seats

Analysts, engineers, and agents — everyone on your team is included. Pricing never depends on headcount.

03

Sized in one call

We quote on call volume, deployment model — managed SaaS or your own AWS — and support level. Annual agreements, no surprises.

Billing and plans

Questions about the bill.

Need a contract tailored to your deployment? Email [email protected].

What counts as a call?

One successful execution against a compiled model. Validation failures, rejected requests, and internal health checks do not count.

How does the free pilot work?

You bring a workbook; we compile it on the managed runtime and validate it against your approved quotes and calculations. No charge and no credit card — the pilot exists so your team can judge the platform on your own models.

How is Enterprise priced?

A flat annual agreement sized to your expected call volume, deployment model, and support level. Overage and cap terms are set out in the order form, and you can set a hard usage cap in the dashboard.

Can we run Eucalyptus in our own AWS account?

Yes. A private execution destination in your AWS account and region is included with Enterprise — calculation traffic, workbooks, compiled models, tokens, and execution history stay under your control.

Do you publish prices?

Not yet. Early deployments are scoped hand-in-hand with the engineering team, so we quote per deployment after one scoping call. Pricing is per call with no per-seat fees.

Start with a pilot · scale in place

Bring a workbook. Leave with an API.

One scoping call covers the pilot, your deployment model, and a quote. Your workbook stays the source of truth.