Changelog
Release history and what changed in each version.
AI Summary Prompts, Provider Presets, and Honest Shareware
The AI summary settings learned three new tricks, audio gained per-source toggles, the processing queue got real per-task progress, and Autorec moved to an honest shareware model.
Editable AI summary prompt
The system prompt that drives every AI summary is now editable. Ask for decisions, action items, risks, deadlines — whatever your team actually reads. The default returns a JSON {title, summary} shape so your library still gets meeting titles automatically; keep that part of the prompt intact and you can rewrite the rest.
Provider presets — one dropdown, twelve LLM backends
Switching between OpenAI, Anthropic, OpenRouter, Groq, Cerebras, SambaNova, Hyperbolic, Ollama, LM Studio, LocalAI, llama.cpp, and vLLM is now a one-click affair. Pick a preset and the base URL plus a sensible default model fill themselves in.
Per-source audio toggles
Mic and desktop audio each get their own toggle, device picker, gain slider, and live level meter. Whole-desktop or specific-apps-only capture, plus two-way / mic-priority ducking so neither side gets buried.
Processing queue with live progress
Every transcription, summary, and compress job lands in a single processing list with per-task progress, status, and cancel controls. Filter by active, completed, or failed. Close the window and the queue keeps working in the background.
Honest shareware licensing
Three free recordings every 24 hours, up to 40 minutes each, with a five-minute heads-up before auto-stop. Pay €20 once and the cap goes away; one key activates up to five devices. No DRM beyond a polite nag — the existing recordings stay yours either way.
Other improvements
- Tray icon turns red while a recording is active so the state is obvious from across the room
- Settings sub-views are now addressable from the screenshot CLI:
--screen settings:ai-prompt,--screen settings:license,--screen library:blocked, and friends - Locale-safe JSON serialization throughout the audio meter pipeline (no more silent fallbacks on
de_DE/fr_FR)
Initial Release
- Automatic meeting detection for Zoom, Microsoft Teams, and Google Meet
- Zero-touch recording with hardware acceleration (NVENC, QSV, VAAPI)
- Window-specific capture — records only the meeting window
- Quality presets: 1080p, 720p, 480p at 15/24/30 fps
- Automatic transcription via whisper.cpp with .txt and .srt output
- Multiple transcription model sizes: tiny, base, small, medium, large
- AI-powered title and summary generation via OpenAI-compatible APIs
- Built-in video library with card and table views
- Search, sort, and play recordings with embedded player
- Transcript viewer with clickable timestamps
- Inline transcript editing
- Trim recordings to remove dead time
- Compress recordings to save disk space
- Processing queue for background tasks
- System tray integration with notifications
- Configurable settings: output directory, filename templates, quality, per-app toggles
- Available for Linux (.AppImage, .deb) and Windows (.exe)