LMNT

LMNT offers a developer-focused text-to-speech API that delivers lifelike, low-latency streaming audio and fast voice cloning from very short recordings. It supports over 24 languages and is optimized for conversational apps, agents, games, and media production. The product includes a free playground for experimentation, straightforward API access, and paid tiers for higher usage and commercial licensing.

• Low-latency streaming stream text-to-speech with 150–200ms latency suitable for conversational agents and games.
• Studio-quality voices access a catalog of polished voices designed for natural prosody and clear delivery.
• Fast voice cloning create a voice clone from as little as a 5‑second sample to maintain consistent branding or character voices.
• Multilingual support produce speech in 24+ languages and switch languages mid-sentence where supported.
• Developer-friendly API and playground start in the free playground, then integrate via the documented REST/streaming API for server or client apps.
• Scalable pricing tiers use the free tier for evaluation and move to paid monthly plans for larger volumes and commercial licensing.

To get started, try the free playground to audition voices and sample cloning, then follow the API documentation to integrate streaming or batch TTS into your application. For production scale, choose a paid plan that matches your monthly character needs and request an enterprise quote for custom concurrency or volume terms.