Skip to main content

Product Roadmap

This page outlines the current and planned features for Knowii Voice AI.

Short-Term Goals​

Recently Delivered ✅​

  • Instant Transcription (Experimental): Transcription runs at natural pauses while you are still talking, so your text lands almost immediately when you stop (shipped in 0.9.0) (#127)
  • Reliable Long Dictations: Long recordings are split at pauses in the speech and stitched back together, so minutes-long dictations no longer come back garbled or empty (shipped in 0.9.0)
  • Parallel Model Loading: The transcription model loads in the background while you record, with a retry if that load fails (shipped in 0.9.0) (#126)
  • File Transcription: Drag & drop audio and video files, or pick them from the tray, and transcribe them into your history as a queue (shipped in 0.8.0) (#153)
  • The transcribe CLI Ships With the App: The command-line tool is bundled in every installer, no separate download (shipped in 0.8.0) (#31, #135)
  • Cancel Shortcut: A dedicated hotkey to discard the current recording (shipped in 0.8.0)
  • Verified Model Downloads: Model downloads are checked against SHA256 checksums and resume correctly after interruptions (shipped in 0.8.0)
  • In-App Log Viewer: Read the app logs from Advanced settings to make bug reports easy (shipped in 0.8.0)
  • Rock-solid Linux Support: Rewritten global shortcuts and native, layout-correct text output on Wayland, reliable operation on GNOME and KDE, automatic keyboard permission setup in .deb/.rpm packages, and graceful handling of desktops without a system tray (shipped in 0.7.0) (#158, #313, #316, #317)
  • Microphone Test: Test your microphone directly from the Audio settings (shipped in 0.7.0) (#304)
  • Clearer Error Feedback: The app now warns you when a recording fails to start or when global shortcuts fail to initialize (shipped in 0.7.0) (#303)
  • macOS Support: Native, signed and notarized builds for Apple Silicon and Intel (shipped in 0.6.0)
  • Linux Support: Native .deb, AppImage, and .rpm packages with Wayland/Hyprland compatibility (shipped in 0.6.0)

Next Releases​

  • Obsidian Integration: First-class support for sending transcriptions straight into your Obsidian vault (#310)
  • Code-signed Windows & Linux binaries (#62): macOS builds are already signed and notarized; Windows and Linux builds are not signed yet
  • Mute/Pause Media While Recording: Disable noise while recording to get the best possible transcriptions (#108)
  • Home screen (#52)
  • Auto-submit mode (#53)
  • Prevent sleep while recording (#109)
  • Smart AI-enabled Post-Processing: Apply custom transformations to transcriptions: remove filler words, fix commonly misrecognized words, use AI. Configure AI-enabled pipelines, detect applications/context and go from raw transcriptions to specialized outputs (#24, #18, #19, #20, #49, #124)
  • Hybrid Push-to-Talk (#139)
  • Shortcut to paste last transcription (#141)
  • Voice Trigger Words/Phrases: Keyboard-less mode where a trigger word or phrase initiates recording and transcription (#132)
  • Portable Version: Uncompress. Run. Transcribe. Everything stored in one folder (#147)
  • History UI/UX Improvements (#155, #170, #119, #98)
  • Better trim silences (#110)
  • Minimize to tray setting (#36)
  • Additional Transcription Models: Support more transcription models (local and remote) (#144, #181, #21, #5)
  • Transcription improvements (#146, #123, #125, #129)
  • Improve audio recordings quality (#166, #142)
  • Deeper Internationalization: Deeper support for translations. Different keyboard shortcuts that transcribe in different languages (#136, #138, #122, #33)
  • CUDA Support: GPU acceleration for NVIDIA GPUs using CUDA for even faster transcription (#174)
  • In-App feedback and error reporting (#105, #106)
  • API: Integrate Knowii Voice AI with other applications and systems programmatically (#143). Scripting and automation are already possible today through the transcribe CLI, which ships inside every installer since 0.8.0 (#47, #131)
  • Usage Statistics: Get statistics about your usage
  • Allow selecting a different folder for models (#40)
  • Allow selecting a different folder for the history (#115, #116, #117, #118)
  • Easily add and use custom sound files

Long-Term Vision​

  • Voice Commands & Action System: Use your voice to trigger concrete actions (#26)
  • Plugins System: Extensibility through plugins (#26)
  • Cross-Device Command & Control: Use the app on one device to command other devices (#25, #7, #137)
  • Context Analysis: Understand what you're doing to generate contextual transcriptions (#16)
  • Speaker Diarization: Identify and label different speakers (#68)
  • Cloud Sync (Optional): Optionally sync history across devices (opt-in) (#23)
  • Themes: Custom color themes and UI customization (#22)
  • Mobile Companion App: Android/iOS app for on-the-go transcription (#6)
  • Reprocess past recordings (#130)
  • Custom Hotkeys: Additional customizable shortcuts (#39)

Future Considerations 🔮​

  • Screenshots + AI analysis: Take screenshots with a shortcut, and get help from AI (#217)
  • Meeting Assistant Mode: Record and transcribe meetings with summaries
  • Translation: Real-time translation between languages
  • Voice Synthesis: Text-to-speech with your voice
  • Custom Model Training: Train models on your voice for better accuracy
  • Collaboration: Share transcriptions with team members
  • Auto-Punctuation Enhancement: Improved automatic punctuation
  • Markdown Support: Direct markdown formatting in transcriptions
  • Code Mode: Optimized transcription for programming

How to Request Features​

Have an idea for Knowii Voice AI? I'd love to hear it!

  1. Check the roadmap above to see if it's already planned
  2. Search existing feature requests on GitHub
  3. Create a new feature request if it doesn't exist, create a discussion, or send me an e-mail

Release Cycle​

I ship releases regularly.

Check the Release Notes to see what's new in each version.

Stay Updated​

To stay informed about new releases and features:

Roadmap subject to change based on development priorities and community feedback.