We shrank our steerable small TTS to 11.84 MB. WER is unchanged, UTMOS drops ~0.1 [p]

WhissleAI released SteerTTS Edge, an edge variant of its SteerTTS v1 steerable English text-to-speech model, at 11.84 MB for everything an app must ship — weights, voices, text front end and Python runtime. The company reports word error rate is unchanged from the original model, while the UTMOS naturalness score drops by roughly 0.1.

A while back I posted SteerTTS v1, a small English TTS you steer with numbers https://huggingface.co/WhissleAI/steer-tts . Today I am releasing the edge variant: https://huggingface.co/WhissleAI/steer-tts-edge Size: everything an app has to ship weights, voices, text front end and Python runtime