Hume AI launches 'Octave', a new breakthrough TTS system with human-like emotional depth

Hume AI has launched Octave, a new AI text-to-speech system that leverages large language model technology to produce speech with human-like emotional depth and contextual awareness.

Unlike traditional TTS systems, Octave adjusts tone, rhythm, and cadence based on text meaning, allowing it to convey emotions such as urgency or sarcasm without explicit instruction. Users can customize voices by providing descriptive prompts, specifying traits like accent and emotional tone.

Octave's capabilities make it ideal for applications like audiobooks, virtual assistants, and gaming. It also emerges as a strong alternative to other AI-generated voice services like ElevenLabs, as shown in a blind comparison study where 180 human raters preferred Octave’s output over ElevenLabs in audio quality, naturalness, and alignment with desired voice descriptions.