> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mayaresearch.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Voices

> Two voices, both speaking every language.

| id       |        |
| :------- | :----- |
| `Ananya` | female |
| `Arjun`  | male   |

Those are the only valid values.

<Warning>
  **Case matters.** The model interpolates whatever it is given straight into
  its prompt, so `ananya` is not the same request as `Ananya` — it produces
  audibly different audio. Anything outside the two above is rejected.
</Warning>

## How much audio a clip is

Roughly **800 characters per minute**, measured across both voices. So 10,000
characters is about 12.5 minutes.

| Characters |  ≈ Audio |
| ---------: | -------: |
|      1,000 |  1.2 min |
|     10,000 | 12.5 min |
|    100,000 |  2.1 hrs |