Why patient audio should not go to a cloud AI
If you dictate your own notes, the recording usually contains names, dates of birth, and clinical details. Pasting that into a general-purpose AI chatbot or a consumer transcription app means handing patient audio to a third party whose retention policies you cannot see or control.
FastScribe takes a narrower approach: transcription runs on our own servers, the audio file is deleted immediately, and only the text is kept so you can copy it into your notes. We keep your text, never your sound.
To be clear about what we are not: FastScribe does not claim medical-grade accuracy and holds no regulatory certification. Treat the transcript as a fast first draft, and review every note yourself before it goes anywhere near a patient record. Give it the same proofreading pass you would give a typed draft.
A dictation workflow that fits between patients
Dictate into whatever recorder you already use, a phone voice memo app or a handheld dictaphone, then upload the finished file. FastScribe transcribes uploaded recordings only, so there is nothing to install and no microphone permission to grant in your browser.
Single-speaker dictation is a comfortable case for our transcription engine: one voice, close to the microphone, no crosstalk. Clinical vocabulary still deserves a careful read-through, especially drug names and dosages, which is another reason the transcript should be your draft rather than your final note.
Export as plain TXT to paste straight into your notes, or SRT and VTT if the recording accompanies a video, such as a teaching case. Every transcript includes timestamps, so you can jump back to the exact moment you dictated a finding and check it against the audio before the file is deleted.
When one free file is not enough
A free account raises that to thirty minutes and one hundred megabytes per file. Pro, at twelve dollars a month, handles files up to two gigabytes and five hours, adds DOCX export for letters and reports, batch uploads for a whole day of dictation at once, a priority queue, unlimited files under fair use, and history kept forever. Weigh that deliberately if your transcripts contain patient details, since audio is still deleted the moment your transcript is ready either way.
Frequently asked questions
Is FastScribe certified for medical or legal use?
No. FastScribe makes no claim of medical-grade accuracy and holds no regulatory or clinical certification. It produces a draft transcript that you, the clinician, must review and correct before using it in any patient record.
Does patient audio ever leave your servers?
Transcription runs using our own transcription engine, so your recording is never forwarded to a third-party AI service. The audio file itself is deleted the moment your transcript is ready.
Can I dictate directly into FastScribe during a consultation?
No. FastScribe transcribes uploaded files only, so record your dictation on your phone or dictaphone first, then upload the finished recording whenever you have a free moment between patients.
How long can a dictation file be?
Your free first file can run up to ten minutes and fifty megabytes with no signup. A free account raises that to thirty minutes and one hundred megabytes, and Pro accepts files up to two gigabytes and five hours long.