# Plans and pricing (/docs/pricing)

> How OmniDimension voice AI plans work, what each plan includes, and how usage is billed.

OmniDimension bills voice AI usage per minute, with a monthly plan that
already includes a block of minutes at a set per-minute rate. You can use
any combination of supported speech-to-text, LLM, and text-to-speech
models on any plan.

  For the latest prices, always check the live pricing page at
  [omnidim.io/pricing](https://omnidim.io/pricing).

## How billing works [#how-billing-works]

* Each plan's monthly price already covers its included minutes at the
  listed per-minute rate. In other words, monthly price divided by the
  per-minute rate is roughly the number of included minutes.
* When you go past the included minutes, extra usage is charged at the
  plan's extra-usage rate.
* Phone number rental and telephony are billed separately from your plan
  (see [Numbers Shop](/docs/dashboard-guides/numbers-shop)).
* Call concurrency (channels) is billed separately too, from your wallet
  balance (see below).
* You can [top up credits](https://omnidim.io/billing) with a custom amount
  at any time from your dashboard.

## Concurrency and channels [#concurrency-and-channels]

Concurrency is how many calls your agents can run at the same time. It
applies to your whole organization, not to a single phone number, so every
agent on every number draws from the same shared pool.

* Your plan includes a base number of channels.
* You can buy more channels from
  [Billing → Telephony → Add channels](https://omnidim.io/billing?tab=numbers).
  Each additional channel is $6.74 per month, paid from your wallet
  balance, and renews monthly. You can release a purchased channel at any
  time.
* Total concurrency is your included channels plus purchased channels. For
  example, 10 included plus 5 purchased gives 15 concurrent calls.

To size concurrency, use: channels = (calls per hour x average call
minutes) / 60. Round up and add a buffer.

## Voice AI plans [#voice-ai-plans]

| Plan            | Monthly            | Per-minute rate        | Included minutes | Extra usage | Knowledge base | Best for                      |
| --------------- | ------------------ | ---------------------- | ---------------- | ----------- | -------------- | ----------------------------- |
| Starter         | $15                | $0.084                 | \~179            | +$0.009/min | 5 MB           | Quick experiments and testing |
| Jump Starter    | $30                | $0.076                 | \~395            | +$0.008/min | 10 MB          | Building and sharing demos    |
| Early Deployers | $40 (10% off: $36) | $0.068                 | \~588            | +$0.006/min | 50 MB          | POCs with a live agent        |
| Growth          | $200               | $0.056                 | \~3,571          | +$0.005/min | 100 MB         | Scaling usage after a POC     |
| Enterprise      | Custom             | As low as $0.04 (₹3.5) | Custom           | Custom      | Custom         | Large-scale production        |

Enterprise adds agent training from recordings, dedicated support,
volume-based discounts, and a custom SLA. To discuss it,
[contact sales](https://omnidim.io/contact-us?reason=general).

## Free trial [#free-trial]

New accounts include 12 minutes of free voice AI usage so you can try the
platform before choosing a plan.

## Flexible model selection [#flexible-model-selection]

Every plan supports any combination of supported models. You are not
locked into a "basic" or "premium" tier: pick the STT, LLM, and TTS
providers that fit each agent, and your usage is billed at your plan's
per-minute rate. See [Voices and languages](/docs/dashboard-guides/voices-and-languages)
for the available models.

## White-label and agencies [#white-label-and-agencies]

If you want to resell voice AI under your own brand, see
[OmniRelay](/docs/omnirelay), which has its own agency plans and per-client
billing.