Blog

Adding unlimited text generation, additional audio formats, sampling rate control, and text normalization to Arcana.

As of May 2025, Rime has completed its SOC 2 Type 2 audit

Our ultra-realistic Arcana and Mist models are now live on SignalWire.

Rime's newest spoken language model is the most realistic you've ever heard.

Rime's new French and German voices enable you to create more authentic, multilingual experiences with our ultra-realistic voice options.​

Rimecaster is a new open source training model to make voice AI sound more human

Rime's flagship TTS model, Mist v2, now supports Pig Latin!

Rime: Authentic AI Voices for Real-time Conversations at Scale

Does your text-to-speech model pronounce your words correctly? With Rime, you can be sure.