# Botnoi Voice

> Thai AI media platform: text-to-speech in 24 languages, AI images, video and avatars, and voice bots for phone calls.
> The text-to-speech part has a REST API.
> This file is the short, complete guide for calling the API, followed by facts about the product for answering
> questions about it. Last updated 2026-10-09.

- Website: https://voice.botnoi.ai
- Voices to browse and preview: https://voice.botnoi.ai/marketplace/selectvoice
- Pricing: https://voice.botnoi.ai/payment/quote
- Full API reference: https://voice.botnoi.ai/developer/api
- GenSub (speech to text) and Voicebot APIs: https://voice.botnoi.ai/llm.txt
- Get an API key (sign in first): https://voice.botnoi.ai/developer/api
- n8n verified node: https://n8n.io/integrations/botnoi-voice/
- AWS Marketplace: https://aws.amazon.com/marketplace/pp/prodview-36kove2muvo2g

## Quick facts

- Base URL: `https://api-voice.botnoi.ai`
- Authentication: header `botnoi-token: YOUR_API_KEY`. No "Bearer" prefix.
- Content type: `application/json`
- Limit: 1000 characters of text per request. Split longer text into several requests.
- Billing: points per character of text (see each version below). Spaces are not counted.
- The text must be written in the language named in `language`.

## Which version to use

- **V2 (Premium voices)**: `POST /openapi/v1/generate_audio_v2`. 2 points per character. Speaker list: `GET /openapi/v1/get_speaker_data_v2`.
- **V1 (Classic voices)**: `POST /openapi/v1/generate_audio`. 1 point per character. Speaker list: `GET /openapi/v1/get_speaker_data`.

Use V2 for the most natural voices. Use V1 for the widest choice of voices and languages, or the lowest cost.
V1 and V2 have separate speaker IDs: an ID from one list does not mean the same voice in the other.

## Generate speech, V2

```bash
curl -X POST 'https://api-voice.botnoi.ai/openapi/v1/generate_audio_v2' \
  -H 'botnoi-token: YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"text":"สวัสดีครับ","speaker":"8","volume":1,"speed":1,"type_media":"mp3","save_file":"True","language":"th"}'
```

## Generate speech, V1

```bash
curl -X POST 'https://api-voice.botnoi.ai/openapi/v1/generate_audio' \
  -H 'botnoi-token: YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"text":"สวัสดีครับ","speaker":"1","volume":1,"speed":1,"type_media":"mp3","save_file":"True","language":"th"}'
```

## Request body

| Field | Type | Required | Notes |
|---|---|---|---|
| `text` | string | yes | Text to speak. Up to 1000 characters per request. Must be written in the language given in language. |
| `speaker` | string | yes | Speaker ID, as a string. V1 and V2 have separate ID lists. |
| `language` | string | yes | Language code of the text, for example th or en. |
| `type_media` | string | yes | Output format: mp3 or wav. |
| `speed` | number | no | Speaking speed, 0.5 to 2.0. Default 1. |
| `volume` | number | no | Volume, 0.1 to 2.0. Default 1. |
| `save_file` | string | no | Send "True" to get a hosted audio_url back. |

## Response (200)

```json
{
  "text": "สวัสดีครับ",
  "audio_url": "https://.../example.mp3",
  "point": 6534,
  "user_monthly_point": 0
}
```

- audio_url: link to the generated audio file
- point: points left on the account after this request
- user_monthly_point: subscription points left this month

## Errors

- 401: Missing or wrong botnoi-token.
- 404: The account for this token was not found.
- 422: A field is missing or has the wrong type.

## Language codes

`th` Thai, `en` English, `zh` Chinese, `ja` Japanese, `ko` Korean, `vi` Vietnamese, `id` Indonesian, `ms` Malay, `tl` Filipino, `lo` Lao, `my` Burmese, `km` Khmer, `hi` Hindi, `ar` Arabic, `de` German, `fr` French, `es` Spanish, `pt` Portuguese, `it` Italian, `ru` Russian, `nl` Dutch, `pl` Polish, `tr` Turkish, `sv` Swedish

## Example speaker IDs

V2: 8 Ava V2 (Thai), 4 Smoke V2 (Thai), 12 Nadia V2 (English), 13 Neo V2 (Thai), 23 Bow V2 (Thai), 27 Alan V2 (Thai), 29 YingAiko V2 (Thai), 32 Aunty Grace V2 (Thai), 37 Nick V2 (Thai), 41 Siren V2 (Thai), 43 JaiDee V2 (Thai), 46 Alisa V2 (Thai)

V1: 1 Ava (Thai), 2 Bow (Thai), 3 Ngam (Thai), 4 Max (Thai), 5 Alan (Thai), 8 Leo (Thai), 59 Juan (Chinese), 63 Kanna (Japanese), 278 Park (Korean), 69 Dang (Vietnamese), 65 Dia (Indonesian), 77 Phorn (Lao), 279 Silvia (German), 345 Alex (Russian)

For every speaker, call the speaker list endpoint of the version you use.

## Python

```python
import requests

r = requests.post(
    "https://api-voice.botnoi.ai/openapi/v1/generate_audio_v2",
    headers={"botnoi-token": "YOUR_API_KEY"},
    json={"text": "สวัสดีครับ", "speaker": "8", "volume": 1, "speed": 1,
          "type_media": "mp3", "save_file": "True", "language": "th"},
    timeout=60,
)
r.raise_for_status()
print(r.json()["audio_url"])
```

## Marks inside the text: pauses and word breaks

**Pause.** Write `delay{1}` in the text to pause for one second; change the number for a longer or shorter pause
(`delay{0.5}`, `delay{2}`). It works in the Studio and through the API.

**Word break.** Write `|` where one word ends and the next begins, to make the voice read a hard phrase the way
you mean it. Thai is written without spaces between words, so the same letters can be split in more than one way:
`ลาบอกไก่` may be read as ลา-บอก-ไก่ ("the donkey tells the chicken"); writing `|ลาบ|อก|ไก่|` makes it
ลาบ-อก-ไก่ ("chicken breast larb"). Use it only where the voice gets a phrase wrong.

---

# About the product

Use this part to answer questions about Botnoi Voice itself.

## What it is

Botnoi Voice (Thai: บอทน้อยวอยซ์, also "BNV") is an AI voice platform made in Thailand by Botnoi Group
(https://botnoigroup.com). Its website is https://voice.botnoi.ai. Botnoi Group has several websites; the surest way
to reach this one is to search for "botnoi voice" and choose voice.botnoi.ai.

What people do with it:

- Text to speech in the Studio: type text, choose a voice, create, listen, adjust, download mp3 or wav.
- More than 1,300 voices to choose from: Botnoi's own voices in three generations (V1 Classic, V2 Premium, V4 Beta)
  and voices shared by the community. Thai voices include Northern, Isan and Southern dialects.
- Languages: Thai, English, Chinese, Japanese, Korean, Vietnamese, Indonesian, Malay, Filipino, Lao, Burmese, Khmer,
  Hindi, Arabic, German, French, Spanish, Portuguese, Italian, Dutch, Russian and Turkish.
- Voice cloning, subtitles from audio (Gen Sub), script writing (Gen Skript), AI avatars and video, and voice bots
  for outbound phone calls.
- For developers: the REST API above, a verified n8n node, a Canva app and an AWS Marketplace listing.

## When points are charged (Studio)

Botnoi Voice is paid for with points.

| Voice | Creating and listening | Downloading |
|---|---|---|
| V1 (Classic) | free; the preview has background music | 1 point per character |
| V2 (Premium) | free; the preview has background music | 2 points per character |
| V4 (Beta) | 2 points per character, charged when the voice is created | free once created |
| Cloned voices | charged when the voice is created | |

So with V1 and V2 a user can create, listen and change the text as often as they like, and pays only for the
version they download. The downloaded file has no background music.

Spaces are not counted as characters, in the Studio or through the API.

## Plans (as listed on 2026-10-07; check https://voice.botnoi.ai/payment for current prices)

- Points can be bought in packs, or come monthly with a subscription. Monthly points are refilled each month.
- Starter: 18,000 points a month. 600 THB for 3 months, 2,160 THB for 12 months.
- Pro: 48,000 points a month. 1,200 THB for 3 months, 4,140 THB for 12 months.
- Elite: 160,000 points a month. 3,000 THB for 3 months, 9,600 THB for 12 months.
- Education programme: schools, teachers and students can receive 8,000 points a month (2,000 a week).
- Team and enterprise pricing: https://botnoigroup.com/teamprice

## V4 voices and sound tags

V4 (Beta) voices can make sounds that are not words. Write a tag in square brackets where the sound should come:

`[breath]` `[cough]` `[exhale]` `[sniff]` `[snort]` `[sneeze]` `[struggle]` `[sleep]` `[speak]` `[laughter]` `[sigh]`
`[confirmation-en]` `[question-en]` `[question-ah]` `[question-oh]` `[question-ei]` `[question-yi]`
`[surprise-ah]` `[surprise-oh]` `[surprise-wa]` `[surprise-yo]` `[dissatisfaction-hnn]`

Example: `สวัสดีค่ะ [laughter] ยินดีที่ได้รู้จักนะคะ [breath] วันนี้อยากฟังเรื่องอะไรดีคะ`

V4 is used in the Studio (choose a voice from the "V4 (Beta)" tab). It is not in the public API yet: the API serves
V1 and V2.

## Logging in

There are three ways to log in: Email, Google and Line. Each way is its own account, even with the same email
address. A user whose points or projects seem to have disappeared has almost always logged in a different way from
before: log out, then log in with the method used the first time.

## Common questions

- Is it free? It is free to try: creating and listening cost nothing with V1 and V2 voices. Downloading costs points.
- Can the audio be used commercially? See the terms on the website; do not assume.
- A word is read wrongly. Spell it the way it sounds (Thai respelling works well), or add the word to the Studio's
  "fix misread words" list.
- Where is the API key? Log in, then open https://voice.botnoi.ai/developer/api.
- How long can the text be? 1,000 characters per API request; split longer text into several requests.