Plans & billing
Callable bills on a metered model: you pick a monthly plan, then pay-as-you-go for the usage you run on top of it.
How billing works
Usage is metered every 60 seconds while a call is running, not once when it ends. Two things follow from that. A long call shows its cost as it happens rather than appearing at the end, and calls have no time limit — a caller and an agent can talk for as long as the conversation lasts, because cost is bounded by a meter that runs alongside the call rather than a stopwatch that cuts people off mid-sentence. Set Max call length on an agent only if you actively want a cap.
What a call is made of
Four things are metered on every voice call, and their relative weight surprises most people. Measured on a real 66-second international call: telephony 47%, the language model 31%, text-to-speech 21%, speech-to-text 1%.
The language model is usually the biggest lever you control. The whole prompt is re-sent every turn, so a large model on a chatty call adds up quickly; choosing a smaller one is a dropdown in the agent builder and often changes nothing a caller would notice. Telephony dominates on international destinations — the same call domestically costs far less.
Your bill has two parts:
- A monthly plan. This sets your baseline — including a usage budget and the limits on how many agents, teams, and users you can create.
- Pay-as-you-go usage. As your agents run, their usage (LLM spend plus voice spend) is metered against your plan's budget.
This means a light month costs less than a heavy one, while your plan keeps the ceilings and features you signed up for.
Free trial
New accounts start on a 7-day free trial — build 1 agent and try it with about $1 of included AI usage, no card required. The trial uses the platform's standard voices (the standard engine). Premium voices (Fish Audio), buying phone numbers, running outbound campaigns, and agent testing/evals unlock once you move to a paid footing — Pay-as-you-go or a paid plan. After 7 days the trial automatically becomes Pay-as-you-go, where you only pay for the usage you run.
The usage meter
Your dashboard shows a usage meter that tracks your metered spend against your plan's budget. Usage accrues from the work your agents do — the LLM generating replies and the voice pipeline (speech-to-text and text-to-speech) handling calls and chats.
Check the meter to see how much of your budget you've consumed so far and how much headroom remains for the rest of the billing period.
Plan limits
Your plan caps how many of each of these you can create:
- Agents — your voice and chat agents.
- Agent Teams — multi-agent teams that hand off to each other.
- Users — teammates you invite into your workspace.
When you reach a limit, upgrade your plan to raise it. See the Pricing page for the limits on each plan.
Paying with Stripe Checkout
Paid plans are purchased through Stripe Checkout.
- Go to the Settings area and open the Billing tab (owner-only).
- Choose the plan you want.
- Complete payment in Stripe Checkout.
Once payment goes through, your plan — including its budget and limits — is applied to your workspace.
Next steps
- Team — invite teammates, set roles, and find the owner-only Billing tab.
- Integrations — connect your own LLM, voice, and transcription keys.
- Co-pilot — ask the in-dashboard assistant about your current plan and spend.