A working draft, not a certified record
FastScribe produces a machine-generated draft, and machine transcription makes mistakes, especially with crosstalk, legal terms of art, and proper names. Treat the output as a starting point: read it against the recording, fix errors, and note anything unclear before anyone in the firm relies on it.
The transcript comes back split by voice, each one labelled Speaker 1, Speaker 2 and so on for you to rename. Treat those labels as a drafting aid and nothing more: the software separates voices by their sound, it does not identify people, and an automated label is not evidence of who spoke. Check every attribution that matters against the recording during your review pass. If you need a certified transcript for filing or evidentiary use, engage a certified court reporter; FastScribe makes no claim of legal, evidentiary, or court-admissible standing.
Keeping sensitive recordings off third-party services
Recordings tied to a matter should not wander through outside AI services, and with FastScribe they do not. Everything runs on our own servers, so the audio you upload is processed in-house rather than being forwarded to another provider.
Once transcription finishes, the audio itself is deleted the moment your transcript is ready. Only the text is kept, so you can come back later to reread the draft or export it again without the underlying recording sitting on someone else's disk.
From a single statement to a whole matter file
One file a week is free with no signup, capped at ten minutes and fifty megabytes, which is enough to test a short recorded statement before committing to anything. A free account raises that to thirty minutes and one hundred megabytes for longer recordings.
When a matter produces hours of audio, Pro at twelve dollars a month accepts files up to two gigabytes and five hours, adds batch uploads for working through a folder of recordings, DOCX export for drafts you circulate inside the firm, a priority queue, history kept forever, and unlimited files under fair use. Everyone can export TXT for reading and markup, or SRT and VTT if the recording is video and you want captions aligned to playback.
Frequently asked questions
Does FastScribe tell me who is speaking in the recording?
It separates the voices and labels them Speaker 1, Speaker 2 and so on, and you rename them yourself. It cannot identify a person from audio, and for legal work the distinction matters: an automated label is a drafting aid, not evidence of who spoke. Verify every attribution against the recording using the timestamps.
Can I file the transcript with a court or use it as evidence?
No. FastScribe output is an uncertified machine-generated draft with no legal or evidentiary standing. It is meant for internal review and preparation; when you need an official record, use a certified court reporter.
What happens to a client recording after I upload it?
The audio file is deleted from our servers the moment your transcript is ready. Only the transcript text is kept, so you can revisit and export the draft without the recording being retained.
Can FastScribe transcribe a video recording of a statement?
Yes, you can upload video files as well as audio. If the recording lives on an online service, download the file to your machine first, then upload it; FastScribe only works with files you upload directly.