01 · TWO USEFUL MODES
Clean when you want flow.
Verbatim when every word matters.
Clean
Transcription followed by a narrow cleanup pass. It removes common fillers, fixes mechanics, and resolves obvious self-corrections while preserving wording, tone, and meaning.
Verbatim
Direct transcription without the extra cleanup model. It is faster and better suited to quotations, exact phrasing, and source material you plan to review yourself.
02 · ONE THOUGHT, TWO RESULTS
The useful difference is small.
Illustrative example of the intended mode behavior:
“Um, send the launch notes Tuesday—actually Wednesday—and, uh, keep the pricing section short.”
“Send the launch notes Wednesday, and keep the pricing section short.”
“Um, send the launch notes Tuesday—actually Wednesday—and, uh, keep the pricing section short.”
Clean removes speech artifacts and honors the correction. It does not decide what the launch notes should say. Verbatim keeps the direct transcript so the source remains visible.
03 · WHERE AI DICTATION ENDS
A transcript is not an email, brief, or plan.
Sometimes you want your words cleaned. Sometimes you want the substance reshaped into finished work. Combining those actions by default makes the product unpredictable: you cannot tell whether the model corrected a sentence or changed the idea.
AudioArcher keeps Transforms explicit. The original and Clean versions remain intact while a polished email, task list, script, or custom workflow is saved as a separate Output.
04 · AI DICTATION ON A MAC
Use the mode where you are already writing.
- 1Choose Clean or Verbatim.
Your preference applies to the next dictation.
- 2Put the cursor in any field.
ChatGPT, Claude, Gmail, Slack, Notes, or another Mac app.
- 3Hold Fn and speak.
Release when finished. The text returns to the field and remains on your clipboard.
05 · COMMON QUESTIONS
AI dictation FAQ
What is AI dictation?
AI dictation converts speech into text and can apply a restrained cleanup pass for fillers, punctuation, capitalization, and obvious self-corrections. It should preserve the speaker’s wording, tone, facts, and meaning.
How is AI dictation different from speech-to-text?
Speech-to-text produces a transcript. AI dictation may add a separate cleanup step so naturally spoken language is ready to paste into everyday work. AudioArcher also offers Verbatim mode when you want direct transcription without that cleanup step.
Does Clean mode rewrite what I say?
It should not. AudioArcher’s Clean mode is instructed not to rephrase, embellish, or summarize. A deliberate Transform is the separate action for turning a note into a different format such as an email, task list, or brief.
When should I use Verbatim mode?
Use Verbatim for quotations, exact phrasing, technical language, interview material, or any situation where preserving the direct transcript matters more than removing speech artifacts.
Can AI dictation work in any Mac app?
Yes. AudioArcher for Mac listens while you hold Fn, then pastes the completed text into the active field and copies it to your clipboard as a fallback.