TTS-Romanian: 720 Hours of Speech from 456 Speakers

TTS-Romanian contains 267,410 speech samples totaling 720 hours of audio derived from Romanian audiobooks. Released by datadriven-company and updated in early 2026, the collection features 456 unique speakers with an average DNSMOS quality score of 3.84.

Use Cases

• Training text-to-speech (TTS) models using the mp3 and text pairs
• Developing automatic speech recognition (ASR) systems for the Romanian language
• Speaker identification or verification tasks utilizing the 456 unique speaker IDs

Strengths

• 720 hours of total duration
• 456 unique speakers
• Average DNSMOS of 3.84

Limitations

• Audiobook source material may result in a formal prosody that differs from natural conversational speech

Provenance

Source

CartiaAudio.eu

Collection Method

Derived from Romanian audiobooks

Freshness

Last updated February 2026

Geography

Romania