← Back to Changelog

Voice generation gains live playback and speech tags

Create voice with Gemini, ElevenLabs, and xAI using richer speech tags, live waveform playback, karaoke timing, and batch previews.

Released in v0.9.33.

Voice mode now supports a broader provider lineup, including Gemini TTS and ElevenLabs alongside xAI voice generation. Hylope also adds a shared speech-tag vocabulary, inline tag insertion, tag profiles, and custom speech cues so performance direction can stay inside the prompt.

The playback experience is richer after generation starts. Streaming voice results use a continuous waveform-style player, batch jobs reveal results in place, audio cards keep their identity across Gallery and input slots, and xAI results can sync karaoke highlighting from character-level timestamps.