v0.1 (beta)

Reference

Plans & limits

Beta / Academic / Pro / Enterprise tiers, concurrency, retention.

TurbineX bills on tokens, not wall-clock compute time. A token is the unit of cost for one OpenFAST job; the cost is derived from the simulation's length and time step, server-side, and is plan-dependent. Plans differ in concurrency, simulation length caps, retention, and discount on the token formula.

Plan tiers

Three tiers are on offer, listed by intended audience rather than price — see the pricing page for the current figures:

  • Academic — discounted token rate for credentialed academic accounts; same operational caps as Pro, with a metered monthly allowance. Eligibility is verified before the plan starts.
  • Pro — for working consultants and small engineering teams. Standard concurrency, max runtime per job, and artifact retention, with a 0.8× discount on the token formula.
  • Enterprise — bespoke caps, dedicated worker pool option, SSO and SAML, 0.5× discount on the token formula, contracted SLA. Speak to us directly.

A fourth tier, Beta, is closed: it was the evaluator programme and is not open to new accounts. Evaluators already on it keep Pro's operational caps at complimentary billing (1.0× on the formula below) under the terms they accepted, which is why it still appears in the discount list and on their Billing tab.

The token formula

One token is one standard simulation: TMax = 660 s at the default time step, i.e. about 150,000 time-steps. A job is billed in proportion to its time-step count, so a short run costs a fraction of a token and a longer or finer-time-step run costs more:

tokens = (time_steps ÷ 150,000) × plan_discount

where time_steps = TMax ÷ DT and plan_discount = 1.0 (Beta), 0.8 (Academic / Pro), or 0.5 (Enterprise), with a 0.01-token floor for very short runs. The Run tab estimates this before submit from the configured simulation length and time step; the actual debit is computed the same way from the job's real TMax and DT once it finishes. Failed jobs (worker error, not user-caused) are not billed.

The Zeph assistant's language-model usage is metered separately and is not counted against your compute token allowance; it shows as its own line on the Billing tab.

Caps and concurrency

  • Concurrency — maximum number of jobs your org can have running in parallel. Excess submissions queue.
  • Max runtime per job — worker kills a job at this wall-clock limit. Long turbulent campaigns can hit this if the simulation length × time step is unusually large.
  • Artifact retention — how long .outb outputs, plots, and reports stay downloadable before the cleaner sweeps them.
  • Max input size — ZIP upload limit for custom turbines.

The exact numbers per tier are listed on the public pricing page and in your Billing tab; we don't freeze them here because they shift as the platform matures.

Monthly spend and token caps

On top of the plan, every org can set two optional ceilings on the Billing tab of the Profile page:

  • Spend cap — a hard limit on monthly checkout charges. Enforced at checkout: you can't buy a top-up that would push the period over the cap.
  • Token cap — a hard limit on tokens consumed per month. Enforced at job submit: the platform returns 402 Payment Required if the new batch's estimated cost would exceed the cap.

Email alerts fire at 50 %, 80 %, and 100 % of either cap, to the org owner. Caps are useful for keeping an exploratory beta budget from running away on a long parametric.

Top-up ladder

When you run out of tokens before the period reset, top up from the Billing tab. Standard ladder: 50 / 200 / 1,000 tokens per top-up at the current per-token price. The top-up increments your current-period balance and rolls over to the next period if unused. The customer portal (Lemon Squeezy) also exposes invoices, payment methods, and plan changes.

What unverified accounts can do

Until you verify your email address the account sits in demo mode:

  • Single Simulation studies against demo turbines — allowed, metered.
  • Custom turbine creation / ZIP upload — blocked (403).
  • Certification and Parametric submissions — blocked (403).
  • Billing / Usage pages — blocked (403). The Notes page is read-only.

Verify the email to unlock everything. The verification button sits in the Demo-mode banner on every page.

Reading your bill

The Profile → Billing tab shows:

  • Current plan label, current period (start → end), and the period reset date.
  • Current token balance (remaining + used this period).
  • Active alerts (low-balance warnings, cap-threshold hits).
  • Links to the customer portal (manage plan, invoices) and the top-up flow.

The Usage page (sidebar, verified-only) gives a more detailed breakdown: per-day token spend, recent batches with cost attribution, and an exportable CSV.

Changing plan

From the Billing tab, click Manage Plan — the Lemon Squeezy portal opens. Plan changes apply at the start of the next billing period; tokens you've already paid for carry over.

Next steps

See Organization & roles for the per-seat permissions, or the public pricing page for the current numbers per tier.