Nari Labs · 2025–26 · dialogue
Dia
A 1.6B dialogue model that generates multi-speaker script audio, including nonverbals. Apache-2.0. Sample unavailable on this bench.
Known for
Two people in one pass, with [laughs] still intact.
Sample unavailable. No redistributable clip was on file. Listen on the official site rather than a fake waveform.
- Approach
- Spoken dialogue
- Year / era
- 2025
- License class
- open
- Cloning
- No
- Streaming
- No
- Emotion controls
- Yes
- Languages (tags)
- en
- Official
- Source