Developer
$0 forever

Evaluate real agent runs on your own work before anyone signs anything.

  • 20 runs per month
  • 1 seat
  • All four agent domains
  • Run history & step traces in dashboard
  • Community support
Start Free
Scale
$1,990 / month

For departments standardizing agent workflows across several teams.

  • 20 seats
  • 2,500 runs per month
  • Priority run queue
  • Audit logs
  • Dedicated success engineer
Get Started
Enterprise
Custom

For organizations with procurement, network, and compliance requirements.

  • Unlimited seats
  • Custom run volume
  • SSO / SCIM provisioning
  • VPC peering
  • DPA & custom terms
  • 99.99% contractual SLA
  • Dedicated support team
  • Invoicing / PO
Talk to Sales

All paid plans: 2 months free with annual billing. Prices in USD.

Compare plans

Every tier runs on the same agent runtime, the same model routing, and the same zero-retention architecture. The differences are volume, controls, and support.

Feature Developer Team Scale Enterprise
Runs / month 20 500 2,500 Custom volume
Seats 1 5 20 Unlimited
Agent domains All 4 All 4 All 4 All 4
API access
Parallel step execution
Priority run queue
Slack & GitHub integrations
Webhooks
SSO (SAML / OIDC)
SCIM provisioning
Audit logs
Zero data retention
VPC peering
Support Community Email Email + success engineer Dedicated team, named contacts
Uptime SLA 99.99% contractual
Invoicing / PO Card Card or annual invoice Invoicing & PO

All plans run on the same infrastructure. Measured platform uptime over the trailing 90 days is ≥99.98% — see the live status page. The 99.99% SLA is a contractual commitment with service credits, available on Enterprise agreements.

How usage works

Runs are the only usage unit on the platform. We deliberately do not meter tokens — the unit you budget for should be the unit you get value from.

One task, one run

A run is one assigned task carried to one completed deliverable: a reviewed PR with inline comments, a due-diligence memo, a content draft, a sourced research summary. Everything inside — planning, intermediate steps, model calls across providers — is included in that single run.

Retries are free. If a step fails and the agent recovers, or a run fails for a platform reason, you are not billed again. A run only counts when it produces a deliverable.

Overage, priced by domain

Past your monthly allotment, runs continue at per-run overage rates, billed monthly in arrears. Rates vary by agent domain because execution depth varies:

  • Content $0.90 / run
  • Coding $1.10 / run
  • Research $1.30 / run
  • Strategy / Reasoning $1.50 / run

You get an alert at 80% of allotment, and you can set a hard cap so runs queue instead of incurring overage.

Runs don't roll over

Allotments reset on your billing date and unused runs don't accumulate. We provision execution capacity against each plan's allotment, which is how we keep latency predictable — banked runs would mean banked capacity nobody is holding.

If you're consistently under or over your allotment, moving tiers mid-cycle is prorated to the day. Most teams check the usage dashboard once a month and adjust from there.

What would this save your team?

Plug in your team's review workload. The model uses the 73% first-pass reduction we measure across production deployments — adjust the inputs to match your reality.

Hours saved / month
Dollars saved / month
Suggested plan
Payback multiple
Estimate based on measured 73% average first-pass reduction across production deployments. Your workload mix will move this number — run the live demo on a representative task to calibrate.

Not sure yet? Watch a run first.

The live demo replays complete agent runs — PR review, due diligence, contract review — step by step, with the same traces you'd see in the dashboard. No signup required.

Watch the Live Demo →

Pricing FAQ

What counts as a run?

A run is one assigned task carried to one completed deliverable — a reviewed pull request, a due-diligence memo, a content draft, a research summary with citations. Everything inside the run is included: task planning, intermediate steps, model calls across providers, and retries. Retries never consume additional runs, and runs that fail for platform reasons are not billed. If a run produces nothing usable, you shouldn't pay for it, and you don't.

What happens when I hit my monthly run limit?

Nothing abrupt. You get a notification at 80% of your allotment. Past 100%, runs continue at overage rates starting at $0.90 per run depending on agent domain, billed monthly in arrears. If you'd rather never see an overage line item, set a hard cap in the dashboard — runs past the cap queue until your next cycle or until someone with billing permissions raises the cap.

How does proration work when I upgrade or add seats?

Upgrades take effect immediately. We prorate the price difference to the day for the remainder of your billing cycle, and your run allotment switches to the new tier at the moment of upgrade — not at the next cycle. Downgrades take effect at the start of the next billing cycle, so you keep what you paid for. Seat changes inside a tier work the same way.

Is there an annual discount?

Yes. Annual billing on Team and Scale includes 2 months free — you pay for 10 months and get 12. Annual plans are invoiced upfront. Enterprise agreements are typically annual by default, with multi-year terms available during negotiation.

How do trials work?

Team comes with a 14-day trial: full Team features, all four agent domains, Slack and GitHub integrations included, no credit card required. When the trial ends, you can subscribe to keep going or drop to the Developer tier — your workflow configurations and integration settings are kept either way (there's no customer content to migrate, because we never retain any). Scale and Enterprise evaluations are run as structured pilots — talk to us to scope one.

How does security review and procurement work?

Usually faster than you'd expect, because there's less to review: FuturOne runs a zero-retention architecture, so customer data never persists beyond the request lifecycle. SOC 2 Type I is complete and Type II is in its observation period — the report is available under NDA. A standard GDPR DPA is available to all paid plans; Enterprise agreements include a negotiated DPA, security questionnaire support, and custom terms. Teams typically clear enterprise compliance review in about two weeks. Full details, including our subprocessor list, are on the security page.

Can I bring my own model API keys?

No, and it's a deliberate design decision rather than a missing feature. FuturOne agents route individual workflow steps across frontier models with automatic retries, failover, and parallel execution — that only works if we own the full execution path. Our zero-retention guarantee also depends on every model call running under our enterprise zero-retention terms with providers; a customer-supplied key would put your data outside that boundary. FuturOne is an agent platform, not a model proxy: you pay for completed deliverables, not pass-through tokens, so there's no per-token margin a BYO key would save you.

Start with 20 free runs

No card, no sales call. Upgrade when the runs are doing real work.

Start Free Talk to Sales