Faster note capture Saving a note is now a single keystroke: just press Return to save, and Shift+Return to start a new line. Prefer the old way? ⌘+Return still saves too. This one came straight from a user request — thank you for the feedback, and keep it coming.

NEW: Audio files • Drop in mp3, m4a, m4b, flac, wav, or aiff and take timestamped notes the same way you do for videos • Podcast listeners, audiobook readers, music students, language learners — VidNotes now works for you too • Audio metadata (artist, album) shows in the player when present NEW: Mermaid diagram export • Export your knowledge graph as a Mermaid diagram (.md) for pasting into docs, Notion, Obsidian, or GitHub • Plus three other graph export formats (Obsidian wikilinks, Logseq blocks, GraphML) that were already in the code but didn't have a button IMPROVED: Settings discoverability • Model pickers now point to the Storage section so you can find the cleanup tools IMPROVED: Reliability • Internal hygiene around dormant feature surfaces, smoke tests • Capture latency instrumentation for ongoing optimization Thanks to everyone who sent feedback on 1.0.2 — keep it coming.

1.0.2 — multilingual transcription, faster note-taking NEW: Multilingual transcription • Pick your language from 15+ presets — Spanish, Mandarin, Japanese, Korean, Hindi, Urdu, Arabic, French, German, Portuguese, Russian, Vietnamese, Indonesian, Malay, and more • Or enter any ISO 639-1 code for languages not in the list • Auto-detect remains the default — no change if you're transcribing English NEW: Faster, smoother note capture • Cmd+Return now saves notes near-instantly — no more pause between keystrokes • Trash and edit icons appear on hover, so cleaning up notes doesn't need a right-click • Notes still anchor to the exact second you pressed Cmd+Return NEW: Playlist sorting • Sort any playlist by Title (A→Z or Z→A), Date Added, or Duration with one click • Imported a 50-video course? Bulk sort beats dragging videos one-by-one • "Manual" mode preserves your custom drag order NEW: See your tier at a glance • A colored tier badge in the toolbar shows whether you're on Free, Scholar, or Researcher • Settings → Account is now the first thing in Settings, with Restore Purchases and offer-code troubleshooting • Redeemed an offer code but don't see the upgrade? Click Restore Purchases — it usually catches up immediately IMPROVED: Knowledge graph • Manually merge duplicate entities ("OpenAI" + "OpenAI Inc" → one entity) with a guided merge sheet • Same-type entities surface first to avoid accidentally merging a person with a company IMPROVED: YouTube playback • Videos with embedding restrictions now show a clean fallback with a "Watch on YouTube" link, instead of a grey overlay • Better error messages when a video can't play in-app IMPROVED: Reliability • AI providers fall back gracefully when one hits a rate limit — clearer messages so you know which provider switched • Optimistic UI means notes appear instantly even on slower disks IMPROVED: Storage management • New "Reset" section in Settings → Storage lets you wipe all downloaded transcription or AI models in one click • Use it to start fresh, reclaim disk, or recover from a corrupted download IMPROVED: Privacy framing across the app • Settings, paywall, and pricing copy all surface what's actually local: transcription, AI briefs, chat, knowledge graph — all run on-device by default • Cloud routing (OpenAI / Anthropic / Gemini) remains opt-in if you bring your own API key

1.0.1 • Fixed YouTube playback that showed a blank screen for users on the released build. Videos now load correctly when added by URL. • Improved error reporting when a video has embedding restrictions.