01

Dictate into a draft, not the final message

On Mac and Windows, Wispr Flow can dictate into an active text field: hold its hotkey, speak, then release to transcribe and paste. Start in a draft or note rather than a live Slack, WhatsApp or email composer. That leaves room to correct the substance before an Enter key, shortcut or hurried click can send it.

Speak in complete thoughts and pause between sections. Do not rely on voice input for exact URLs, account numbers, names or quoted wording; paste or check those against the source after transcription.

Wispr Flow: starting a dictation

02

Finish the wording before formatting it

Read the transcript once for meaning. Correct names, numbers and missing words; remove spoken asides; then decide what the reader needs first. A channel update may need a decision and actions, while a Word document may need headings and a longer explanation.

Keep this editing pass separate from destination conversion. Otherwise it becomes hard to tell whether a changed sentence came from dictation cleanup or from the formatting step.

  • Confirm names, figures and dates
  • Add links by copying their real destinations
  • Put the decision or request first
  • Split long speech into short paragraphs
  • Approve the wording before changing its format
03

Translate the approved draft for its destination

Paste the finished draft into Pasteful, choose the destination and inspect both Preview and Text. WhatsApp needs sendable message markers and readable fallbacks; Slack's visual composer differs from API mrkdwn; Word should receive editable headings, lists and links rather than the dictation app's visual styling.

If a heading, nested list, table or labelled link cannot survive in the destination, rewrite that structure as a short lead-in or labelled lines. Copy only after the preview still communicates the same meaning without depending on colour or font size.

Open the local formatter · Choose a destination-specific tool · Check destination limitations

04

Know where the privacy boundary sits

Wispr Flow's official documentation says voice transcription requires an internet connection. Pasteful's formatting and deterministic cleanup run in the browser tab, but that later local step does not make the earlier dictation local or undo data already handled by the dictation service.

For confidential work, check your organisation's approval and Wispr Flow's current privacy terms before dictating. If networked transcription is not permitted, type the draft or use an approved offline input method, then use Pasteful only for the final destination conversion.

Wispr Flow: how Flow works · Pasteful's local-processing boundary

Format your own text

Paste it into Pasteful. It stays in this tab.

Open Pasteful