Pricing
Vio is usage-based: you pay per minute of AI conversation, and the exact ₹ cost of every call is shown in your call log. The per-minute rate depends on the engine you choose for each agent — a cheaper pipeline engine or a premium realtime engine — so you control the cost per agent.
Self-serve billing is coming soon. During Beta there are no plans or checkout to set up — usage is metered automatically and the cost is visible on every call in your call log. Prices below are indicative and may change.
What you pay for
Each agent runs on one of two engines, and the engine sets the per-minute AI rate:
| Engine | How it works | Indicative price |
|---|---|---|
| Pipeline | Separate transcriber + language model + voice, chained together. Fast, natural, and the most cost-effective — the right default for most agents. | ~₹2.5–3 / min |
| Realtime | A single speech-to-speech model that hears and speaks directly, for the lowest latency and most fluid turn-taking. | ~₹15 / min |
You set the engine per agent on the Engines & voices tab, so a high-value sales agent can run on Realtime while a routine FAQ agent runs on the cheaper Pipeline engine.
Telephony is billed separately
Vio’s per-minute price covers the AI — listening, thinking, and speaking. It does not include the phone minutes (the actual PSTN/mobile call), which are billed by your own carrier or Twilio on your telephony account.
Think of it as two lines on the bill:
- AI (Vio) — metered per minute by engine, shown per call in your call log.
- Phone line (your carrier / Twilio) — metered by your telephony provider at their rates.
If you bring your own carrier over a SIP trunk, the phone minutes stay entirely on your existing telecom bill.
How to control cost
- Pick the right engine per agent. Most agents do not need Realtime — Pipeline is roughly 5× cheaper per minute. See Engines & voices.
- Choose an efficient voice. Different voices carry different costs; the Engines & voices tab shows your options, including Virtuo Speech (Indian) voices tuned for Indian callers.
- Keep calls tight. A focused persona and prompt that gets to the point ends calls sooner, and you pay per minute.
- Watch the numbers. The Analytics page shows average duration and cost so you can spot expensive patterns.
FAQ
Q. How am I charged today? Usage is metered per minute based on each agent’s engine, and the cost of every real call is shown in your call log. Self-serve plans and checkout are coming soon; during Beta there is nothing to configure.
Q. Why is the Realtime engine more expensive? Realtime uses a single speech-to-speech model that processes audio directly for the lowest latency and most natural turn-taking. That capability costs more per minute than the Pipeline engine, which chains cheaper components. Use Realtime where fluid conversation really matters, and Pipeline everywhere else.
Q. Is the phone number’s cost included in Vio’s per-minute price? No. Vio’s per-minute price is for the AI only. The phone minutes and any number rental are billed separately by your carrier or Twilio. See SIP trunks and Numbers.
Q. Where do I see exactly what a call cost? Open the call in your call log — each real call shows its total ₹ cost with a per-component breakdown. In-browser test calls are free and are not logged.