Frequently asked questions

We support MP3, WAV, M4A, AAC, FLAC, AIFF, OGG, Opus, AMR and WMA audio, and MP4, MOV, MPEG, WebM, MKV, AVI, WMV and 3GP video. Files can be up to 100 MB without a subscription and 200 MB with Weekly or Premium. You can also paste a YouTube link or a direct URL to a file hosted online.

Processing is typically a fraction of the original length — often in the range of a few minutes for a short meeting. Quick, Standard, and Premium tiers trade speed for accuracy. Exact times depend on audio length, complexity, and current load.

Transcribr supports a wide set of languages including English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Chinese, Japanese, Korean, Arabic, and many more. Language detection is automatic, or you can specify a language for better accuracy.

We aim for industry-leading accuracy on clear audio — often 95–99% depending on tier and conditions. Results vary with audio quality, background noise, accents, and technical terminology. Premium modes prioritize accuracy; Quick is built for speed.

Quick prioritizes turnaround for drafts and rough notes. Standard balances speed and accuracy for most meetings. Premium adds higher accuracy and stronger speaker identification when every word and speaker label matters.

Yes. Transcripts are fully editable. Tap any section to adjust text, and rename default speaker labels (Speaker 1, Speaker 2, etc.) to real names so exports stay clear for your team.

Every account gets 10 free minutes a day in total, to use on quick, standard or premium in any mix, for up to 3 transcripts a day. Premium on free minutes includes speaker labels. Free minutes reset at 00:00 UTC, and you don't need a card.

Credits let you transcribe beyond your free minutes without a subscription. For uploads and recordings, one credit buys 1 hour of premium, 2 hours of standard or 4 hours of quick. Packs cost $4.99 for 3 credits, $9.99 for 10, $21.99 for 25 and $44.99 for 60. Your free minutes are always used before your credits.

You pay once, with no subscription. Packs stack into one balance, and each pack lasts 12 months from the day you buy it. When you have credits from more than one pack, the oldest are used first. Credits are sold on the web only, not in the mobile app stores.

A YouTube link uses your free minutes like any other file. With credits, it costs 1 credit an hour on quick and standard, and 2 credits an hour on premium.

Quick and standard are unlimited. Premium is covered under fair use, up to about 25 hours a month. Past that, premium pauses until your plan renews, or you can pay for it with credits, while quick and standard stay unlimited.

Yes. Cancel Premium or Weekly anytime from your account settings. You'll keep access until the end of the current billing period, then your account goes back to 10 free minutes a day.

Yes. Audio and transcripts are encrypted in transit and at rest. We don't sell your content. You own your files and can delete data from your account when you no longer need it. See our Privacy Policy for full details.

Yes. When you turn backup on, notes and recordings are encrypted in your browser with AES-GCM, using a key derived from your passphrase with PBKDF2. Only the encrypted result is stored, and the passphrase never leaves your device, so nobody here can decrypt it. Backup is off until you enable it, and you can keep the encrypted copy as a file instead.

Export to PDF, TXT, SRT, VTT, and CSV. You can choose plain text only, or include speakers and timestamps depending on your workflow — notes, captions, or archives.

Yes. Conversion, compression, clip extraction, and the batch queue run in your browser. The source and completed files stay on your device and are not uploaded to Transcribr. Available output formats depend on what your browser can encode.

Yes. You choose where the output goes before conversion starts, and the audio is written to that file as it is produced instead of being assembled in memory, so a long recording will not exhaust the tab. You also see an estimated output size before starting, and you can play the result before saving it.

No. Live Notes saves recognised text only. Audio is created only if you explicitly enable local recording before starting; that recording stays in this browser and is never uploaded automatically. This is true whether you capture your microphone or a browser tab.

Recognition works wherever the browser exposes on-device speech recognition for your chosen language, so support varies by browser, operating system and language — Transcribr checks the actual capability rather than blocking all mobile devices. Capturing another tab’s audio additionally needs Chrome or Edge on desktop, and the option is hidden where it cannot work.

Yes, on Chrome or Edge on desktop. Choose “A browser tab” or “Both” before starting a session and pick the tab to share, making sure “Also share tab audio” is ticked in that dialog. Microphone capture alone only hears the room, so on headphones the other participants are not picked up at all. Tab audio is still recognised on your device — nothing is uploaded.

Yes. A recording made during Live Notes stays on your device, and when you want an accurate transcript you can send that same recording for cloud transcription in one step. The on-device notes stay exactly where they are.

They are stored in this browser using local device storage. Clearing site data can remove them, so you can export an AES-256 encrypted backup of notes and optional recordings and restore it later with your password. If you would rather not manage a file, encrypted cloud backup keeps a copy off your device that only your passphrase can open.

Three tools, each with its own screen. The caption editor plays the media beside the text so you can fix wording and timing against what was actually said. The clip editor draws a waveform and exports the range you pick. The redaction tool finds sensitive values in a transcript or text file and lets you review each one before it is applied. Everything runs in your browser and the source files are not uploaded to Transcribr.

Yes, and you should. Every match — a name, email, phone number, link, payment card, bank account, national identifier, API key, address, postcode, date, or a value announced by a label such as “Policy ID” — is listed for you to accept or reject on its own, with the replacements highlighted in the text. Detectors err toward flagging too much, because a false positive you untick costs you a click and a missed detail costs you the leak.

Yes. On Pro you can open a finished transcript directly in Private Studio — its subtitles in the caption editor, or its text in the redaction tool. Both run in your browser. The caption editor supports undo and redo and refuses to export invalid timings; redacted text exports as TXT, Markdown, DOCX, PDF, HTML or JSON.