Miso One

What is Miso One?

Miso One is a voice model page for Miso Labs’ Miso TTS 8B. It focuses on expressive English text-to-speech with conversational style, emotional delivery, voice continuation, and low-latency research. You can try a hosted demo, browse voice models, and check pricing for higher character limits and voice credits. The product also highlights open weights, local inference, and prompt-audio based voice cloning experiments for developers. It is aimed at people who want to evaluate speech quality, latency, and how well the model fits creator workflows or voice-agent setups.

Key features

• Expressive English text-to-speech with natural conversational delivery.
• Supports voice continuation using prompt audio context.
• Offers a hosted demo for quick quality testing.
• Provides open weights and local inference access.
• Includes voice cloning and voice design workflows.
• Shows latency and hardware considerations for developers.