ona

Core Workflows

Export and Settings

Export finished work as subtitles, JSON, plain text, or Markdown, copy text to the clipboard, and quickly find Dashboard, Diagnostics, Backup & Restore, Storage Management, Cloud Sync, Automation, API Server, LLM Service, Shortcuts, Voice Typing, and Message Center entry points.

Ask AI

This deployment has not enabled protected guide Q&A.

This page covers the final stage of the workflow: exporting finished transcripts, doing final speaker/version checks, locating the settings areas that matter most, and knowing where support surfaces such as Dashboard, Diagnostics, Backup & Restore, Automation, and notifications live.

Export a finished transcript

Before exporting, open Speaker Review from the transcript header if speaker labels matter for your delivery or review process. For saved non-draft transcripts, open Version Snapshots first if you need to recover selected rows or revert a bulk rewrite.

  1. Click the Export button in the header.
  2. In the Export Transcript modal, enter a Filename.
  3. Choose an Export Directory.
  4. Pick an output format: SubRip (.srt), WebVTT (.vtt), JSON (.json), Plain Text (.txt), or Markdown (.md).
  5. Choose an export mode: Original, Translation, or Bilingual.
  6. Click Export.
  7. Use Copy to Clipboard when you want the same selected mode as plain text without writing a file.

What export gives you

  • Sona writes the transcript to the selected path and format.
  • Markdown (.md) preserves speaker labels as bold labels and keeps the transcript readable in Markdown editors.
  • If translation text exists, you can export translated-only or bilingual output.
  • Copy to Clipboard uses the selected export mode and copies plain text from final segments.
  • Translation and Bilingual only appear when at least one segment already contains translation text.

When to go back to Projects

  • If you want to reopen, rename, move, or sort saved recordings and imports, go back to Projects and Inbox.
  • Project Center is also where project context and Inbox organization live now.

Settings areas organized by category

The Sona desktop settings sidebar is grouped into 5 labeled categories, with a client-side search bar at the top of the modal for instant cross-tab filtering:

1. General

  • Settings > General: theme, app language, font, auto-start on boot, silent launch (start minimized to system tray), update checks, plus Diagnostics and Backup & Restore
  • Settings > Dashboard: global content overview, speaker coverage, and LLM usage trends
  • Settings > About: version info, build channel indicator (dev, ci, nightly, stable), release notes, license, and update actions

2. Recognition & Speech

  • Settings > Input Device: microphone selection, system audio selection, microphone boost, Keep Microphone Active, and mute during recording
  • Settings > Subtitle Settings: floating caption behavior, window styling, window positioning, click-through, and font sizing
  • Settings > Voice Typing: turn on Voice Typing, assign its global shortcut, choose Push to Talk (Hold) or Toggle (Press once), situational presets, and check readiness
  • Settings > Model Settings: Live Record Model, Batch Import Model, CTC forced alignment, diarization sensitivity (Permissive/Balanced/Strict), Batch VAD Segmentation, VAD Buffer Size, and model management
  • Settings > Vocabulary: organized into sub-tabs (Recognition & Correction, Prompts & Context, Speaker Profiles) and unified dictionary (Visual Mode & YAML source)

3. AI & Automation

  • Settings > LLM Service: feature model bindings (local offline GGUF via llama.cpp or cloud LLMs via aimux), reasoning options with thinking separation, and provider credentials
  • Settings > Automation: watched-folder rules that automatically transcribe, polish, translate, and export incoming media into target projects

4. Integration & Sync

  • Settings > Cloud Sync: End-to-End Encrypted (E2EE) cross-device sync vault supporting WebDAV and S3-compatible object storage (AWS S3, Cloudflare R2, MinIO, etc.), Master Password, Emergency Recovery Key, and Conflict Center
  • Settings > API Server: local HTTP API host, port, optional API key, IP allowlist, REST transcription, and WebSocket streaming (includes migration hint to Tools > API Server)
  • Settings > MCP Server: Model Context Protocol bridge configuration, endpoint monitoring, and agent permission toggles (includes migration hint to Tools > MCP Server)

5. System & Advanced

  • Settings > Storage Management: data and models directory overview, hot-reload storage migration (live directory moving with progress modal), pure-Rust audio decoding with optional custom FFmpeg fallback path, disk usage breakdown, and cleanup policies
  • Settings > Shortcuts: live recording, playback, search, project navigation, and editor shortcuts (supports strikethrough Ctrl + Shift + S)

Diagnostics, backup, cloud sync, and message center

  • In Settings > General, use Diagnostics to inspect the local transcription chain, runtime readiness, and packaging environment.
  • In the same page, use Backup & Restore to export or import a light archive of config, project data, light history transcripts and summaries, automation state, and dashboard LLM usage.
  • Light backup archives restore text history and summaries, but not original audio files. Restored entries may reopen for reading and editing without playback.
  • Settings > Cloud Sync is a standalone End-to-End Encrypted (E2EE) incremental sync system. Transcripts and settings are encrypted with your Master Password before leaving your device, with an Emergency Recovery Key for access recovery, and a Conflict Center to arbitrate concurrent edits across devices. Audio recordings remain stored locally on your machine and are never uploaded to the cloud.
  • An entry at the bottom of the Cloud Sync page allows a one-time import of legacy WebDAV backup archives; Sona no longer creates new full remote backup bundles.
  • The Message Center in the header centrally manages batch transcription, download notifications (multi-task queue, resumable downloads, real-time speed monitoring, and cleanup), automation, and LLM jobs with real-time progress and cancellation for batch tasks; interrupted tasks can also be resumed or discarded in its Recovery Center.

If you mainly came for an extension capability

Useful notes

  • This page is intentionally not a full settings manual. It is here to point you at the right next page.
  • If you are troubleshooting rather than exporting, go straight to FAQ and Troubleshooting.
Previous

Projects / Inbox

Next

AI Summary