Pashto Voice Collector پښتو غږ راټولوونکی

Why this exists — دا پروژه ولې؟

Voice assistants, transcription tools and AI models barely understand Pashto — because almost no open voice data exists for our language.

This project crowdsources real Pashto recordings from real speakers, then publishes them as a clean, open dataset that anyone can use to train AI — so the next generation of technology speaks our mother tongue.

How it works — طریقه

1. 1

   Press the button above — the recorder loads instantly, already warmed up in the background.

2. 2

   Pick one of three ways to contribute — read an AI sentence, write your own, or upload a voice note.

3. 3

   Record and submit. Every clip is reviewed, then exported into the open dataset as high-quality audio.

Your privacy — محرمیت

Contributions are anonymous — no names, no accounts. Each speaker gets a random ID. You only share your tribe, gender and age so researchers can balance dialects fairly.