Skip to content

1.14.0

Choose a tag to compare

@github-actions github-actions released this 17 May 07:39
· 316 commits to main since this release

Bragi Canvas 1.14.0.

Changes

  • Added a unified Qwen Voice model entry that routes internally between built-in Qwen TTS, voice reference synthesis, and voice design synthesis.
  • Added Design as a voice source mode. It uses an upstream text node as the voice design prompt while keeping the current node text as the spoken script.
  • Renamed the custom voice source UI to Voice ref and kept it separate from Design.
  • Added DashScope voice design support for Qwen Voice and CosyVoice, including custom voice metadata stored as bragiCustomVoices.
  • Removed the separate user-facing Qwen Flash / Qwen VC / Qwen Instruct model entries in favor of the single Qwen Voice entry.
  • Kept the Qwen built-in voice picker wired to the existing Instruct Flash sample list.
  • Updated MCP generation to preserve voice source parameters such as voiceMode and voiceDesignTextIndex.
  • Bumped the plugin version to 1.14.0.

Release assets

  • manifest.json
  • main.js
  • styles.css