TrainScription

Live meeting transcription that runs entirely on your device. No bot joins your call. No audio uploaded to the cloud. Speaker-labeled, timestamped, and corrected by a vocabulary Brain you train yourself. ────────────────────────────────────────── NO BOT. NO CLOUD. ANY APP. JUST THE WORDS. ────────────────────────────────────────── TrainScription captures meeting audio locally using OpenAI Whisper running via WebAssembly — directly in your browser, with no network requests during transcription. Your audio never leaves your device. Full Desktop mode captures system audio from any app making sound on your machine — not just browser tabs. If your machine can hear it, TrainScription can transcribe it. 0 cloud calls during transcription. 0 external servers processing your audio. 0 bots joining your call. ────────────────────────────────────────── HOW IT WORKS ────────────────────────────────────────── 1. Open your meeting as normal — in a browser tab OR a desktop app. 2. Click Begin Capture. Choose Browser Tab mode to capture a meeting running in Chrome, OR Full Desktop mode to capture any app making sound on your machine. The extension captures two audio streams on independent pipelines — your microphone and the meeting audio — each transcribed on its own channel, so overlapping speech on both sides is captured, not just whichever channel is loudest. A silence gate drops empty chunks per channel before Whisper ever sees them. No server, no API key, no internet connection required during transcription. 3. Watch the transcript build as your meeting runs. Audio is transcribed in windows, so lines appear roughly 20 to 30 seconds behind the conversation — each one speaker-labeled with your configured channel names and a precise timestamp. Close the popup at any time — transcription continues. 4. Highlight misfires to train the Brain. See a word the AI got wrong? Highlight it. A popover appears instantly — assign the correct spelling in one click. That mapping is now permanent. Every future transcript corrects that word automatically, before you ever see it. 5. Sessions are always recoverable. When you end a session, TrainScription finishes transcribing the audio still in its queue before saving, so the closing moments of a conversation are not lost — or stop early and keep what has already been transcribed. Long sessions are automatically segmented into sealed chunks, grouped together. Export any segment or the full session as a .txt or .srt file. Review any past segment to catch and correct misfires after the fact. 6. You bring your AI. TrainScription gives you the clean local transcript — what you do with it is up to you. When you're done, open Recovery and send the transcript to whichever AI platform you already use and trust — with one click. Your instructions and transcript are copied to your clipboard; your trusted AI platform opens ready to receive them. Nothing is sent automatically, and TrainScription never sees the content after it leaves your device. ────────────────────────────────────────── FULL DESKTOP MODE ────────────────────────────────────────── Full Desktop mode captures all system audio — your microphone plus everything your machine is playing. Select your screen from the OS picker, enable "Also share system audio," and TrainScription captures the full dual-channel audio stream the same way it always has. Any desktop application producing audio is transcribed — video conferencing clients, media players, voice chat, recorded webinars — anything making sound on your machine. No browser tab required. The Brain works the same way in both modes — your trained vocabulary corrects misfires automatically, in Full Desktop or Browser Tab sessions alike. ────────────────────────────────────────── AUDIO & VIDEO FILE IMPORT ────────────────────────────────────────── Already have a recording? Import an existing audio or video file — a meeting export, a webinar recording, anything you already have — and run it through the same on-device Whisper pipeline used for live sessions. No live capture required. Your file is read locally in your browser and processed entirely on your device, exactly like a live session. It is never uploaded anywhere, at any point. The Phonetic Brain applies automatically to imported files the same way it does to live sessions — your trained vocabulary corrects misfires before you ever see the transcript. Audio & Video File Import is a Pro feature. ────────────────────────────────────────── BULK EXPORT, BACKUP & RESTORE ────────────────────────────────────────── Export every stored session at once as a single .zip file — no more downloading segments one at a time. Available on Free and Pro. Pro adds Backup Everything: a complete, machine-readable backup of every session and your full Phonetic Brain, packaged as one .zip file. Restore from Backup brings it all back — into a fresh install or a new machine — without losing anything you've already saved. Useful for moving to a new computer, reinstalling the extension, or simply keeping your own copy of everything you've captured. ────────────────────────────────────────── THE PHONETIC BRAIN — THE DIFFERENTIATOR ────────────────────────────────────────── Every AI transcription tool has the same problem: proper names, product names, case names, medication names, campaign terms, company jargon — mangled, every time, in predictable patterns. The Phonetic Brain is the permanent fix. Train it once on a word. It corrects that word on every future transcript automatically, before you ever see the output. It compounds — the more you use it, the more accurate your transcripts become. Four ways to train: · HIGHLIGHT — Select any misheard word during a live session or while reviewing a past segment. A popover appears — assign the correct spelling in under three seconds. · MANUAL ADD — Already know how Whisper mangles a term? Use the Manual Add form — correct spelling plus the variants Whisper produces. · …