РусскийEnglish

Bayalig · an open corpus of the Buryat language

The Buryat language corpus

Bayalig gathers texts and voices of Buryat speakers into one open corpus. Every sentence is reviewed by native-speaker experts — with care and respect for local dialects.

Telegram bot VK bot

No account needed: everything happens inside the bot.

The project in numbers

The same figures as on the statistics page. Recalculated once an hour.

updated 7 days ago

Hours of speech recorded

1.2h

+0.3 h this week

Every recording, including the ones still being reviewed

Participants

7

+2 this week

Accounts in the bots

Sentences approved

84of 213

+12 this week

213 suggested in total

Recordings

96

+18 this week

Audio recordings in the corpus

First milestone: 100 hours of speech

1.2%

1.2 of 100 hours

98.8 h to go

This is just the start: the bar is honest, with no artificial filling.

How to help in three steps

Nothing to install — only the messenger you already have.

  1. 01

    Open the bot

    Telegram or VK, whichever you prefer. You can start in a minute.

  2. 02

    Suggest a sentence or record your voice

    A line from a book, a song or a conversation. Or record your voice — thirty seconds is enough.

  3. 03

    Tell your family

    The more speakers join, the fuller and more accurate the corpus becomes.

The corpus is growing

All statistics

1Corpus growth

The corpus grew by 0.3 hours this month — 1.2 hours in total

Cumulative hours of speech and participants by month

0.00.51.01.50510 0.30.40.60.80.91.2May 2026Jun 2026Jul 2026Aug 2026Sept 2026Oct 2026
Hours of speech Participants

For researchers and partners

The corpus holds texts and audio recordings of the Buryat language — the material speech recognition, synthesis and translation are trained on.

Data is collected in the open, and every sentence is checked by native-speaker experts. Data access and partnerships are discussed individually.

Contact the team
Steppe and mountains in Buryatia — a warm, film-like frame