IndicVoices speech data collection by AI4Bharat (IIT Madras) with Bhashini funding: 7,348 hours from 16,237 speakers in 22 Indian languages, 1,639 hours transcribed at release
$5-$6 Mn contributed by Bhashini to AI4Bharat for data collection per programme of 7,348 hours of audio released
Dated 7 Mar 2024Source checked 21 Oct 2026Not re-read automatically: the site's terms or its robots.txt do not allow itBasis: reported
In the source's own words
“The state-backed Bhashini project has reportedly contributed $5-$6 Mn to AI4Bharat for the purpose of data collection ... Of the total 7,348 hours of audio, a majority (74%) is extempore while the remaining is read (9%) and conversational audio.”
Where it was read
What it is about
- Bhashini
- AI4Bharat
- Audio and speech
Figures it counts in
- $5.5M per contractAudio and speech, cost to produce, per contract
Cited in this guide