Pricing · Verified per-workload benchmarks · Cancel anytime

Pick the tier your fleet actually needs.

Three plans priced against the same verified cost-per-workload figures on the model routing guide. Free trial. No credit card.

Starter
$0/mo

For small teams kicking the tires on a single OpenAI or Anthropic account.

  • 1 API key, 1 provider
  • Email alerts on spend thresholds
  • Daily spend dashboard & reports
  • 30-day spend history
  • Community support
Start free trial →
Free 3-day trial. No credit card.
Scale
$199/mo

For high-volume production agents and teams who need hands-on help from day one.

  • 10+ API keys, every supported provider
  • Autonomous rerouting & spend forecasting
  • Custom thresholds & per-agent budgets
  • Dedicated onboarding & quarterly reviews
  • SSO-ready & audit-log export
  • Named-account support, 4-hour SLA
Talk to sales →
Tell us your fleet size — we'll route you to a sales engineer.
Anchored to verified benchmarks

On a 1M-call/mo, 500-token blended workload, the routing guide's calculator estimates Gemini 2.0 Flash at $50/mo input, GPT-4o Mini at $75/mo, Claude Haiku 4.5 at $400/mo — and that's just input. Pick the tier that matches the volume — and the failure cost — of the workloads you're routing.

See the verified per-workload benchmarks →
Questions?

Most teams we talk to have two or three open questions about pricing, security, and integrations. We collected the eight most common ones here.

Read the FAQ →
See the workflow

Walk through the four-step workflow — quick-add entry, category tagging, daily & weekly rollups, and export — and see which tier each capability ships in.

See the workflow →