Uzbek text to speech
Paste Uzbek text on the website, or send it to @ovozai_app with /tts, pick a voice, and hear it read aloud: announcements, lessons, audiobook chapters, voice-overs for videos. Up to 20,000 characters at a time, typed or from a Word or PDF file. The same OvozAI that does Uzbek speech to text, the other way round.
- 10 minutes free every month
- 8 female and male voices
- Up to 20,000 characters
How it works
From text to audio
- 01
Add the text
On the website, type or paste it, or import a .txt, .doc, .docx or PDF. In the bot, send /tts and then the text: messages and files are joined, 20 to 20,000 characters in all.
- 02
Pick a voice
The website has 8 female and male voices, each with a sample to hear; /tts in the bot offers two of them. You choose for each text.
- 03
Listen, forward, download
The bot sends a voice message you can forward to a group or channel. The website plays it, lets you download it, and keeps your recent ones in a history.
Example
A shop's announcement, read aloud
156 characters of Uzbek sent with /tts, the female voice picked, and a 12-second voice message back. Below, the text and a sample of every voice.
Hurmatli mijozlar! Do'konimiz yangi manzilda ochildi: Toshkent shahri, Amir Temur ko'chasi, 15-uy. Har kuni soat to'qqizdan yigirma ikkigacha sizni kutamiz.- ZephyrFemale · Bright
- LedaFemale · Youthful
- AchernarFemale · Soft
- CallirrhoeFemale · Easy-going
- CharonMale · Informative
- AchirdMale · Friendly
- AlgiebaMale · Smooth
- IapetusMale · Clear
What it handles
What it handles
Uzbek text of any kind, from a one-line notice to a book chapter.
Up to 20,000 characters
Long texts are read in parts and joined into one recording, so a chapter or a lecture script comes back as a single file.
Word, PDF and text files
Import .txt, .doc, .docx or PDF on the website, or send them to the bot. A PDF needs selectable text; for a scan, copy the text out first.
8 voices
Four female and four male voices on the website and the API: bright, youthful, soft, informative and more. The bot's /tts offers two of them.
Voice message or file
In Telegram it arrives as a voice message, ready to forward. On the website, play it, download it, or find it later in the history.
Charged once delivered
Each second of audio takes 6 seconds of your time, only once it reaches you. If making it fails, nothing is charged.
Through the API
POST /v1/audio/speech gives MP3, Opus or WAV, at 400 credits per 1,000 characters of text.
Price
10 minutes free every month, then from 857 UZS a minute
Text to speech uses the same time as transcription: a minute of audio takes 6 minutes of it, so the free 1 hour a month makes 10 minutes of audio. The packs are the same too, with no subscription, and purchased time never expires.
- Free10 minutes1 hour of time · every month
- 10 minutes10,000 UZS$1.491 hour of time · 1,000 UZS a minute
- 30 minutes27,000 UZS$2.993 hours of time · 900 UZS a minute
- Best value70 minutes60,000 UZS$6.497 hours of time · 857 UZS a minute
A 12-second announcement like the one above: the free 10 minutes make 50 of them every month. A 15-minute audiobook chapter takes half of the 3-hour pack (27,000 UZS); the other half reads the next chapter, or transcribes 90 minutes of recordings.
FAQ
Questions
How do I convert Uzbek text to speech?
Open Text to speech on the website, sign in with Google or Telegram, paste the text and pick a voice: the audio plays in the page and downloads. In Telegram, send /tts to @ovozai_app, then the text. Anything from 20 to 20,000 characters works, and 10 minutes are free every month.
Is there a free Uzbek TTS?
Yes. Text to speech uses the same time as transcription, 6 seconds for each second of audio, so the 1 hour free every month makes 10 minutes of speech, with no card. After that, the 1-hour pack (10,000 UZS) makes 10 minutes, and the 7-hour pack comes down to 857 UZS a minute. Purchased time never expires and transcribes recordings too.
Can I turn a Word document or PDF into audio?
Yes. Import a .txt, .doc, .docx or PDF on the website, or send it to the bot after /tts; several files and messages are joined into one text of up to 20,000 characters. A PDF has to contain selectable text: for a scanned book, copy the text out or run it through OCR first. The audio comes back as one recording.
Which voices are there, and what can I use them for?
8 on the website and the API: four female and four male, described as bright, youthful, soft, easy-going, informative, friendly, smooth and clear. You can hear a sample of each on this page; the bot's /tts offers two of them. Tags such as <laugh> or <sigh> in the text add a laugh or a sigh. People use them for announcements, lessons, audiobooks and voice-overs.
Is there an Uzbek text-to-speech API?
Yes. POST /v1/audio/speech on the OvozAI API is OpenAI-compatible and returns MP3, Opus or WAV in any of the 8 voices. It costs 400 credits per 1,000 characters of text, from the API's own credits, and your first key comes with 1,800 of them free. See the speech-to-text API page for the rest.
More from OvozAI
Telegram voice messages to text
Forward a voice message to the bot and read it a few seconds later, in Latin or Cyrillic.
Learn moreUzbek audio to text
Upload a lecture, interview or meeting, hours long, and get every word of it back as text.
Learn moreUzbek speech-to-text API
An OpenAI-compatible API for Uzbek transcription, subtitles and speech, paid in so'm or by card.
Learn moreOvozAI vs ElevenLabs
Made for Uzbek: Latin or Cyrillic, and speech mixed with Russian
OvozAI is 4.8× cheaperRead
Hear your first text
10 minutes free every month. No card, nothing to install.