CHANGELOG.md

1.12.3 (2026-04-29)

Features

• settings: add an auto-paste toggle so users can keep transcripts in history without automatically inserting them (1c47ac3)
• support: add in-app bug and crash report submission with system info and the latest redacted app log excerpt (77a9b26, be90102)

Bug Fixes

• support: harden report validation, stale-submit guards, fallback behavior, and log redaction for private report delivery (b8291e3, 6953429, fa45160)
• ui: restore scrolling for Upload, General, Help, Advanced, and Formatting settings pages after the sidebar layout changes (551fc4e, 10eb66f)

1.12.2 (2026-04-27)

Bug Fixes

• transcription: fix truncation in long recordings and preserve explicit insertion boundaries (4719f3d, 4627145)
• transcription: preserve sentence spacing on insert and use compatible punctuation matching (987d881, bb63e21)
• recording: harden push-to-talk startup, license fallback, and abort handling (2205bc1, d38f8cb, 7328b5c, aa30d5e, 86d2f17)
• windows: reduce Windows crash paths and bundle the WebView2 offline installer for clean first-run installs (4719f3d, 8c58cd2)
• ui: keep sidebar navigation accessible and route settings navigation to the correct tab (1d70be3)
• license: show the Manage License button when a license is already active (4719f3d)
• ai: persist AI formatting across restart by warming cached credentials from secure storage (482dacb)
• updater: remove stale v1 dialog config and use a versioned update marker for post-update UX (ad84a1d)
• autostart: move autostart toggles to backend-owned commands for atomic OS state management (9aa7bdd)

1.12.1 (2026-02-22)

Bug Fixes

• openai: handle token and temperature compatibility for gpt-5 (42300b9)

Refactors

• ai: remove anthropic provider support (2caf98c)

1.12.0 (2026-02-21)

Features

• add sound on recording end with toggle option (12095a3)
• add pill indicator mode dropdown with Never/Always/When Recording options (7dfa707)
• add recording indicator position setting with 6 options and configurable edge offset (e2b71f5)
• macos: add Intel Mac (x86_64) support (96abe0f)
• language-aware AI formatting and stale microphone validation (ccb8f56)
• add multi-provider AI formatting support (OpenAI, Anthropic, Gemini) with curated model lists (a5084a9)
• add pause media during recording feature (b501ca1)
• add crash reporter UI with GitHub issue integration (cbd836a)

Bug Fixes

• windows: hide console window flash when starting recording (99c467d)
• windows: ARM64 crash fix - disable Vulkan GPU, optimize CPU threads (1b91ba7)
• windows: bundle runtime and harden fresh install behavior (a9b3577)
• audio: prevent app hang when audio device doesn't respond to cleanup (1a011c6)
• audio: use platform-specific stream cleanup to prevent Windows hang (816ff98)
• add delay after start sound for Bluetooth headset compatibility (6d4a681)
• ui: remove unwanted white ring border and add consistent ring styling to pill and toast (57ddae1)
• media: improve pause/resume reliability and use osascript playback toggle on macOS (c7487fb)
• isolate custom and openai provider key/config handling (48ee91b)
• slider showing range instead of single value (dba6e7c)
• per-request HTTP clients and media resume on recording errors (9a480ad)

1.11.2 (2025-12-16)

Bug Fixes

• auto-install updates in background and use updateService in AboutSection (76ec247)
• avoid dropping empty password in Windows signing (f60cd8b)
• use -f flag for private key file path in signer (93811c5)
• use pnpm tauri signer in Windows script (a4e6e65)
• use pnpm tauri signer instead of cargo tauri (cf5d085)

Features

• macos: toggle dock icon based on main window visibility (0854459)
• session-aware auto-updates with notification support (cdd7d89)

1.11.1 (2025-12-14)

Bug Fixes

• add debug output and explicit GITHUB_TOKEN export (340ebc9)
• add missing typecheck and requireCommits check from release-it (a4cbd35)
• add skipChecks to bypass release-it token validation (72e5e64)
• add toast window to capabilities and increase pill-toast gap (138c902)
• align OpenAI config flow (#34) (80c2cb6)
• auto-set GITHUB_TOKEN from gh CLI for release-it (9469a0e)
• avoid double log extension (820bb1e)
• defer mic permission prompts until onboarding complete (a51b379)
• ensure audio device listeners clean up (51d5e3f)
• improve AI config auto-selection (311c61e)
• improve OpenAI config UX with Update button and better switch visibility (0a0fa64)
• improve recording feedback with sound, throttle, and duration gates (a0627c1)
• improve toast notification with dynamic width and better error messages (04f90f1)
• license errors, app reset, and tauri config (70238af)
• redact API key logging (e0ceb8d)
• remove unnecessary hooks and additional directories from AI config (ade912b)
• route recording errors via pill toast (3fba56f)
• sync microphone list with system (6de9eaa)
• treat only watchdog timeout as timeout (edc260e)
• use gh CLI for GitHub releases instead of release-it (89a4d8f)
• Windows device ID fallback + license status timeout (27ab86d)

Features

• add --build-only flag to release script for resuming builds (7328a35)
• add --dry-run flag to release script (6778622)
• add clipboard retention preference (ba80d62)
• improve AI model availability and auto-selection (a44bdd9)
• move pill/toast lower and auto-reposition on monitor change (39ab78b)
• queue critical pill events for reliable delivery (2799928)
• simplify pill UI to minimal 3-dot indicator with separate toast window (defc907)

Changelog

1.11.0 (2025-12-14)

Bug Fixes

• add debug output and explicit GITHUB_TOKEN export (340ebc9)
• add skipChecks to bypass release-it token validation (72e5e64)
• add toast window to capabilities and increase pill-toast gap (138c902)
• align OpenAI config flow (#34) (80c2cb6)
• auto-set GITHUB_TOKEN from gh CLI for release-it (9469a0e)
• avoid double log extension (820bb1e)
• defer mic permission prompts until onboarding complete (a51b379)
• ensure audio device listeners clean up (51d5e3f)
• improve AI config auto-selection (311c61e)
• improve OpenAI config UX with Update button and better switch visibility (0a0fa64)
• improve recording feedback with sound, throttle, and duration gates (a0627c1)
• improve toast notification with dynamic width and better error messages (04f90f1)
• license errors, app reset, and tauri config (70238af)
• redact API key logging (e0ceb8d)
• remove unnecessary hooks and additional directories from AI config (ade912b)
• route recording errors via pill toast (3fba56f)
• sync microphone list with system (6de9eaa)
• treat only watchdog timeout as timeout (edc260e)
• Windows device ID fallback + license status timeout (27ab86d)

Features

• add clipboard retention preference (ba80d62)
• improve AI model availability and auto-selection (a44bdd9)
• move pill/toast lower and auto-reposition on monitor change (39ab78b)
• queue critical pill events for reliable delivery (2799928)
• simplify pill UI to minimal 3-dot indicator with separate toast window (defc907)

1.10.0 (2025-10-24)

Bug Fixes

• audio: prevent empty or ultra-short recordings from causing errors; add mode-specific min durations (PTT≥1s, Toggle≥3s); auto-recover from Error→Idle and gate start transitions (97e9fad)
• avoid panic when requesting parakeet sidecar (855a2ed)
• macOS): bundle ffmpeg/ffprobe via platform config and resolve lookup in Contents/MacOS to prevent normalization failures in packaged app; chore(stt/soniox: update async model to stt-async-v3 (17108f6)
• persist Parakeet model selection across restarts (fffba52)
• resolve Parakeet sidecar MLX module loading and build issues (33c3913)
• tauri: scope Parakeet sidecar to macOS builds and improve UX (2f54a53)
• tray: gate selection until onboarding completes; add tests for tray label and selection logic; ui: simplify Upload section copy (0ac219a)

Features

• ✨ update windows ffmpeg (734fddb)
• add Parakeet MLX sidecar integration (ead96e3)
• parakeet: add version-aware v2/v3 download/load/delete with FluidAudio 0.6.1; fix false-success download for v2 by verifying loaded model and cache; autoload selected Parakeet model on startup (bec2e3e)
• replace Python Parakeet with Swift/FluidAudio implementation (10909df)
• tray: unify model selection and keep tray/dashboard in sync (748675e)

1.9.0 (2025-10-12)

Bug Fixes

• ai: restore Groq/Gemini enhancement path and keep OpenAI-compatible config; cache non-openai keys in backend (b2664f8)

Features

• ai/openai): accept any OpenAI-compatible model; remove dead_code attr. refactor(ui,keyring): standardize Tauri args to snake_case for OpenAI test/save. chore(audio: keep mutable bindings in normalizer for future ops. (5167f0c)
• ai: add OpenAI-compatible provider and endpoint normalization; fix pill feedback timing; cleanups (a28030b)
• audio: normalize all inputs to 16k mono s16 + 5s gate & reduce (937bbb3)
• help: include hashed Device ID in ‘Copy System Info’ via new get_device_id Tauri command; hide mic/accessibility permission lines on Windows to avoid confusion (ac80b7b)
• onboarding: add centered model legend (Speed, Accuracy, Size, Recommended) before model list to match dashboard (3257614)
• tray: add Recent Transcriptions copy, Recording Mode submenu, and Check for Updates; sync Dashboard↔Tray mode changes; refresh tray on history changes (7accfdb)

1.8.0 (2025-09-20)

Bug Fixes

• Add support for multi-channel audio recording (4+ channels) (54e2a0e)
• Enable macOS Cmd+Ctrl hotkey combinations (e427cae)
• Extend offline grace period to 90 days for Issue #15 (7f7bef8)
• Implement secure licensing with grace periods (b123c36)
• Improve hotkey system and shorten feedback messages (bc5ee35)
• Improve language selection UX with alphabetical sorting and name-based search (5f2d631)
• Reduce drag-and-drop area height for better UI balance (5a3dd8c)
• Remove redundant audio validation to solve Issue #16 "Audio too quiet" (b977f7b)
• resolve UnwindSafe trait issues in panic prevention tests (f431ee1)
• Restore download progress updates for model downloads (e3c13b8)

Features

• Add audio file upload and transcription support (beed3d0)
• Add audio file upload for transcription with multi-format support (38f0c58)
• add export to JSON functionality with backend implementation (7a18ce1)
• add Help section and share stats functionality (e67dc68)
• add microphone selection with bidirectional sync (1791519)
• Add push-to-talk recording mode (#13) (a6ef9ae)
• Comprehensive hotkey system improvements with physical key support (6938472)
• enhance history page with search and improved UI (8bd2135)
• Enhance HotkeyInput with editing state notification and update onboarding hotkey initialization (43dca35)
• Enhanced license activation with OS metadata for better device identification (c7b52da)
• Improve history card display (d1a0e67)
• improve UI/UX for support links and Quick Tips section (c3f818f)
• optimize transcription updates with append-only system (5852078)
• redesign Advanced section UI/UX (ae5f1ca)
• redesign Formatting page (formerly Enhancements) with improved UI/UX (e006317)
• redesign Models page with improved UI/UX (53f8837)
• redesign overview dashboard with streak tracking and interactive stats (415d730)
• redesign Settings page with modern UI (2d77331)
• split Account section into License and About sections (4a06489)

Performance Improvements

• implement caching layer for settings and license to reduce I/O overhead (391d752)
• optimize transcription flow by reducing delays and improving UX (22e303a)

1.7.0 (2025-08-20)

Bug Fixes

• handle misheard filler words in base enhancement prompt (b2db125)
• hotkey display platform-specific symbols (04dde4b)
• make enhancement prompt comprehensive for all error types (73a1612)
• resolve menu bar updates and AI enhancement issues (6f3b15f)
• restore license expiration dashboard navigation on hotkey press (0cdbfef)
• update enhancement mode descriptions for accuracy (29d2679)

Features

• improve AI enhancement settings discoverability (84dac26)

1.6.2 (2025-08-14)

Bug Fixes

• enable runtime hotkey updates without app restart (f2701f8)

Features

• ✨ add audio validation and enhanced error logging system (d9ecc06)
• enhance logging system with comprehensive diagnostics and monitoring (64fb1c2)
• implement comprehensive production-ready logging system (5c8bee1)
• implement smart Windows installer with automatic GPU detection and fallback (f314952)

BREAKING CHANGES

• Removed apply_pending_shortcut command (was internal only)

1.6.1 (2025-08-14)

Features

• ✨ add audio validation and enhanced error logging system (d9ecc06)
• enhance logging system with comprehensive diagnostics and monitoring (64fb1c2)
• implement comprehensive production-ready logging system (5c8bee1)
• implement smart Windows installer with automatic GPU detection and fallback (f314952)

1.6.0 (2025-08-08)

Features

• ✨ add Windows dual build system with CPU and Vulkan GPU variants (302d2f0)
• ✨ move vulkan feature flag to windows-specific dependency for whisper-rs (f3dae0b)
• ✨ set small English model as recommended for improved language support (f95087f)
• add GPU-aware smart installer with auto-detection and update system (3319d1e)
• add smart GPU detection and installer for Windows with Vulkan support (81ba6ca)
• add Windows dual build release script with CPU and GPU variants (ad05880)
• add Windows GPU support with Vulkan installer and release scripts (6d85863)
• add Windows release script with MSI installer and update artifacts (ed55678)
• implement robust Windows GPU dual-build support (0aa71b6)

BREAKING CHANGES

• Windows now has separate CPU and GPU builds

Changes:

• Add Vulkan GPU support for Windows (works with NVIDIA, AMD, Intel)
• Implement macOS Metal fallback to CPU on initialization failure
• Add VULKAN_SDK pre-flight check in Windows release script
• Enhance README with GPU requirements and troubleshooting
• Improve runtime logging to show active backend

Critical fixes after review:

• Fix macOS Metal early return to ensure consistent logging
• Remove impossible conditional compilation branches
• Track actual GPU usage for accurate backend reporting

The CPU build ensures universal Windows compatibility while the GPU build with Vulkan provides ~2-3x faster transcription for users with compatible GPUs.

1.5.0 (2025-08-06)

Bug Fixes

• 🐛 simplify Windows window flag application by removing redundant error handling (2c04f14)

Features

• ✨ add timeout to recorder thread, improve window handling, and enhance security validations (758231d)
• ✨ refactor: optimize platform detection with static boolean values (204e4f3)
• add advanced settings section with permissions, reset options and diagnostics (56bcc9b)
• add license management and log cleanup functionality (5834ce6)
• add platform-specific settings and advanced configuration UI (18d5e46)
• add Windows release scripts and configuration files for cross-platform builds (e1516de)
• refactor the app with cross-platform permissions and settings UI (930f9a2)

1.4.0 (2025-08-01)

Bug Fixes

• 🐛 test and remove sentry (57be5a3)

Features

• ✨ add AI enhancement integration with API key management and model selection (994f21b)
• ✨ add Gemini provider integration and related tests (1263503)
• ✨ added sentry well (2f1ceb5)
• ✨ fixes (682c267)
• ✨ implement AI enhancement options and settings management (843f830)
• ✨ implement better key storing (e6a5344)
• ✨ improve checker state (876fda7)
• ✨ improve prompts and ui (5cd995d)
• ✨ investigating audio issue (fd70cc9)
• ✨ little fix and improvements (d481603)
• ✨ refactor audio processing and permission handling (bdc7f35)
• ✨ remove debug logging and add proper resampling (1b415a2)
• ✨ remove deprecated Sentry test commands and related documentation (5aab0a7)
• ✨ update ai enchance and general settings (e844915)

1.3.0 (2025-07-22)

Features

• ✨ add automatic update checks and tray menu functionality, suppress blank and non speech voice (c812355)
• ✨ add huge perf improvement and improved audio level meter and silence detection (aa2601a)
• ✨ add new models, keep only 2 permission, add reset app data (3d3dd56)
• ✨ add sentry, fix esc not stopping, add pill tooltip feedback (82c88b1)
• ✨ add translation, make download and read seperate action (3e9e61a)
• ✨ enable GPU acceleration and multi-threading for improved transcription performance (1f3c1b3)

1.2.2 (2025-07-20)

Features

• ✨ fix download updates (6d72f2d)

1.2.1 (2025-07-20)

1.2.0 (2025-07-20)

Features

• ✨ enhance permissions management in OnboardingDesktop, add automation permission checks, and update sidebar to include advanced section (ca56910)

1.1.1 (2025-07-19)

Features

• ✨ add tauri-plugin-macos-permissions-api dependency, enhance model management in App component, and improve accessibility permission handling (fd96e05)
• ✨ fix external link handling in AboutSection, add updater capabilities in default.json (d627ff5)
• ✨ refactor model management integration in App and OnboardingDesktop components, enhance loading state handling, and improve model status response structure in Tauri commands (82fd144)
• ✨ remove modelManagement prop from OnboardingDesktop, update useModelManagement hook for onboarding context, and adjust event handling for model downloads (d0e079c)
• ✨ show loading while verifying downloads (1197069)
• ✨ streamline model download handling in Tauri commands, enhance logging for download progress, and simplify event emissions in useModelManagement hook (462ad1d)

1.1.0 (2025-07-18)

Features

• ✨ fix license cache (435f8a5)
• ✨ remove CI, release, and test workflows from GitHub Actions for project restructuring (c763ae0)
• ✨ reorganize imports in App component, update active section state, enhance AboutSection with app version retrieval, and clear license cache on startup for fresh checks (762e158)
• ✨ replace Twitter icon with XformerlyTwitter in AboutSection and reorganize imports for better structure (97c05c6)
• ✨ update .gitignore to include release files, clean up AboutSection and LicenseContext for improved readability, and fix URL in license commands (1f9d501)
• ✨ update script (ff5ebeb)

1.0.0 (2025-07-18)

Bug Fixes

• 🐛 pill showing (b67c1d1)
• 🐛 silence audio checking after recording (4de6aa1)
• 🐛 sync with recent history (0ccecec)

Features

• ✨ add compact recording status setting, enhance feedback messages in RecordingPill, and implement cancellation handling in transcription process (f3b42ff)
• ✨ add dialog plugin support and enhance model management with improved UI interactions and state handling (9b7d32e)
• ✨ add formatting script to package.json and refine model management with updated model descriptions and UI enhancements (2c2a997)
• ✨ add iOS spin animation, refactor ModelCard component, and update RecordingPill to use IOSSpinner (f30422e)
• ✨ add new dependencies and update configuration for VoiceTypr (34d30c9)
• ✨ adjust compact mode styles in RecordingPill for better visual consistency and update audio wave animation scaling (fb7e4c7)
• ✨ adjust RecordingPill layout for better feedback message display and update window size calculations for improved visibility (22409d4)
• ✨ clean up release-universal.sh by commenting out test command and removing unnecessary blank lines for improved readability (0d39040)
• ✨ dd autostart functionality with settings toggle, update dependencies, and improve UI components for better user experience (da24d37)
• ✨ enhance audio recording features with real-time audio level visualization, implement update checks in AboutSection, and integrate tauri-plugin-updater for seamless application updates (6494a96)
• ✨ enhance audio visualization in AudioWaveAnimation with improved animation logic and state management, and refactor RecordingPill for better feedback handling and cleanup on unmount (971c9ee)
• ✨ enhance transcription management with cleanup settings and history retrieval, improve UI with pill widget toggle, and update tests for new settings (943d398)
• ✨ enhance transcription management with cleanup settings and history retrieval, improve UI with pill widget toggle, and update tests for new settings (9ffa3f2)
• ✨ implement download cancellation feature, enhance download progress management, and improve hotkey input handling (3ea8ab8)
• ✨ implement model sorting and management enhancements, including balanced performance scoring, model deletion, and improved hotkey input handling (bfcbb86)
• ✨ implement transcription deletion feature, enhance AboutSection with update checks and external links, and improve RecentRecordings component with history refresh (be5b9e6)
• ✨ improve UI components and enhance error handling in recording and model management (6d26e9a)
• ✨ improvements (a042c86)
• ✨ init (3089d43)
• ✨ integrate react-error-boundary for enhanced error handling and improve performance with useCallback and useMemo optimizations in App component (cc92f93)
• ✨ make the pill working right (010d96f)
• ✨ nhance RecordingPill with feedback messages for transcription events, improve ESC key handling for recording cancellation, and update GeneralSettings with tips for users (89a9e07)
• ✨ nhance VoiceTypr with new features, including global shortcut support, model management, and improved UI for recording and settings (ca47873)
• ✨ refactor GitHub Actions workflow for improved macOS target handling and streamline dependency installation (c89eaff)
• ✨ remove recording form the window (5d51fae)
• ✨ reorganize imports and enhance transcription handling in App and useRecording hook (5b217a1)
• ✨ update app title to VoiceTypr, add framer-motion and geist dependencies, and remove unused SVG files to streamline project assets (04c41ee)
• ✨ update dependencies in package.json, add styles to pill.html for full height layout, and remove unused components to streamline the codebase (747d8bf)
• ✨ update IOSSpinner component for improved animation timing and adjust RecordingPill layout for better alignment (a4f269b)
• ✨ update window size for better visibility, remove unused download cleanup logic, and streamline cancellation handling (3ed8ac0)