VoiceStudio Releases

This release improves the update experience, fixes model download interruptions, and enhances UI polish.

Jul 27, 2026

v0.4.1 focuses on stability and bug fixes, notably restoring AMD GPU support and improving error messaging across the application.

Jul 26, 2026

This release introduces plan-scoped workspaces and improves the review-fix loop for subagent-driven development. It also refines skills documentation by removing persuasion prose and restructuring content.

Jul 24, 2026

This release includes fixes for various PII regex patterns globally, migration of the CI toolchain to uv, and new features such as batch deanonymization and support for Philippine UMID identification.

Jul 22, 2026

v0.4.0 introduces major Audiobook and Dubbing features, including multi-voice casting, a designed-voice Gallery, and AI integration tools.

Jul 21, 2026

Introduces the SOP engine, WebAssembly plugin host, Git forge channel, and a self-contained desktop app. Includes cost accounting and security hardening.

Jul 16, 2026

This release focuses on dubbing performance, stereo audio restoration, and memory fixes for 16GB machines.

Jul 13, 2026

A memory fix release that addresses crashes on 16GB machines by stopping TTS engine model stacking and releasing idle dictation models. Introduces a comprehensive new Storage and Reset UI.

Jul 12, 2026

Fixes a race condition in backend restart detection that caused false errors and moves the uninstaller functionality into the application UI.

Jul 12, 2026

Improves error honesty and user experience with streaming audio playback, a proper uninstaller script, and more transparent crash reporting.

Jul 12, 2026

Focuses on self-sufficiency by adding one-click installation for IndexTTS-2 and automatic discovery of working Hugging Face endpoints on restricted networks.

Jul 11, 2026

A major polish release that bundles FFmpeg and yt-dlp, compacts the Engines and Models pages, and fixes several UI and backend stability issues.

Jul 11, 2026

Improves quality with per-sentence caching for audiobooks, text normalization, and enhanced dubbing consistency features.

Jul 10, 2026

Cold-start fixes that resolve timeouts during model downloads, restore the Clear History button, and make auto-play audio stoppable.

Jul 9, 2026

UX improvement that makes ASR and LLM engines visible in the Settings picker, previously hidden behind low-discoverability tabs.

Jul 9, 2026

Community fixes that enable voice cloning on mlx-audio, fix macOS microphone permissions, and add an OpenAI-compatible transcription backend.

Jul 8, 2026

Engine selection honor and network fixes that allow selecting specific MLX-Audio models and resolve dead-ends on restricted networks.

Jul 8, 2026

Adds dark/light pairs for Chroma styles, new template functions for encoding and crypto, Pandoc citation support, and sub paths in layouts. Fixes a performance regression.

Jul 6, 2026

A multi-language reliability release that refines the 'Generate N dubs' workflow to translate per language without destroying work, introduces self-documenting backend crash logs, and fixes SOCKS-proxy and WebView2 compatibility issues.

Jul 5, 2026

A listening and reliability release that fixes hallucinated dialogue, invisible timeline boxes on Windows, and ensures the 'Speakers' count is honored across all dubbing paths.

Jul 4, 2026