Mist v3
Text-to-speech API
Rime's conversational text-to-speech model, optimized for real-time dialogue in production voice agents and available via API or self-hosted deployment.
- Consistently sub-100ms time-to-first-byte
- 600+ voices with customizable accent, pace and tone
- 50+ languages with regional dialects and code-switching
- Deterministic pronunciation control for brand names, proper nouns, numbers and IDs
- On-premises and self-hosted deployment for compliance requirements
Built for
Enterprises and voice AI developers running high-volume automated phone interactions, notably food ordering, healthcare, financial services, hospitality and customer support