Dia-1.6B
1.6B dialogue TTS that generates multi-speaker conversations with emotion.
About
A 1.6B text-to-speech model from the two-person South Korean startup Nari Labs, built for dialogue: it generates multi-speaker conversations in one pass with laughter, sighs, and emotional inflection, plus zero-shot voice cloning from seconds of reference audio. Apache 2.0, ~19k GitHub stars, runs on an 8GB GPU.
Current limitation: English only.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Nari Labs two-person South Korean startup; ~19.4k GitHub stars on nari-labs/dia; Apache 2.0 confirmed on the repo and multiple secondary sources; English-only generation is a documented limitation; repo creation date 2025-04-19 used as release date. License: Apache 2.0 — full commercial use permitted..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 7.0/10
- Openness · license + docs
- 9.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Nari Labs two-person South Korean startup; ~19.4k GitHub stars on nari-labs/dia
- · Apache 2.0 confirmed on the repo and multiple secondary sources
- · English-only generation is a documented limitation; repo creation date 2025-04-19 used as release date