FAQ

winWhisper turns speech into text two ways. Switch between them from the pill in the desktop app — Live is the default.

Both count usage by the length of audio you actually speak, at the same rate — against your free monthly minutes or credits (on Pro, batch dictation is unlimited). Separately, AI Rewrite reshapes any text you select in any app, and Write as Me lets your AI tools write in your voice.

Modes apply to Classic transcription only — they decide how your recording is processed into text:

Live transcription shows your words as you speak; to reshape live (or any) text afterwards, use Improve / AI Rewrite.

AI Rewrite reshapes text you select in any app — press the rewrite hotkey or the ✎ button on the pill, then pick a style and (optionally) a format. It also powers the Improve button on Live and Classic results.

• Styles change how it reads — fix grammar, concise, formal, friendly, translate, plus your own
• Formats change the shape — email, message, letter, bullet points, meeting notes, and more
• Choose what happens to the result: replace the selection, add below, copy, or review first

Output stays in the original language unless you ask it to translate. Styles and formats are fully editable in Settings. (Styles & formats are an AI Rewrite feature — distinct from Classic's modes above.)

Write as Me learns your personal writing voice so winWhisper — and AI agents you connect through its built-in MCP server — can draft text that sounds like you. It is optional and off by default.

• When you opt in, the samples you contribute (your dictations and/or rewrites) are stored on our servers, encrypted, as your writing-style corpus
• It learns from a large sample, so keep dictating (or paste your writing) until there's enough to build an accurate profile — and rebuild it as you add more
• From it we build a style profile — a compact description of your tone and habits (text traits, not your voice audio). Write in several languages? It captures each and replies in the language you write in
• Connect any MCP-compatible AI tool (Claude, Cursor, …) with your API key and ask it to "write as me"
• View, add, or delete your samples and profiles — or turn it off entirely — anytime from the dashboard

Your corpus is used only to provide this feature to you; it is never used to train shared models or shared with anyone else.

This applies to Classic transcription. You can pick between two models to balance speed, accuracy, and language coverage:

Live transcription uses a separate real-time engine (90+ languages) and doesn't use these model settings.

In Classic transcription, the Clean, Rewrite, and User modes add a second AI step on top of the raw transcription — first your speech is transcribed, then a language model refines it to your chosen mode.

This two-step process is what produces polished, instruction-tailored text, and it's reflected in the credits used. Raw mode and Live transcription do not use this extra step.

A few easy ways to translate:

No, your audio and transcriptions are never permanently stored on our servers. Your privacy is our priority:

• Audio (Live and Classic) is processed in real time and immediately discarded
• Transcribed text and AI Rewrite results are only returned to your device
• We only retain minimal metadata (timestamps, usage statistics) for service operation

The one exception is Write as Me (optional, off by default): if you opt in, the samples you choose to contribute are stored on our servers, encrypted, to build your personal writing-style profile. You can view or delete them, or turn the feature off, anytime. See our privacy policy for details.

A single Classic transcription and a Live session are each capped at 5 minutes, which keeps processing fast and reliable.

For longer content, break it into shorter segments — or just keep dictating, since Live auto-stops after a pause and you can continue.

Credits are deducted based on your actual audio duration in seconds, converted to minutes. For example:

• 30 seconds = 0.5 minutes of credits
• 2 minutes 15 seconds = 2.25 minutes of credits

Processing time doesn't affect credit usage — only your audio length matters. (AI Rewrite and Write as Me are billed by the text they process.) On Pro, dictation and AI Rewrite aren't metered at all — credits only apply to Live minutes beyond the included 300.

Pro ($4.99/month or $39/year) gives you:

• Unlimited Classic dictation and AI Rewrite (fair use — normal personal dictation is never metered)
• 300 minutes of Live real-time transcription per month, then credits as top-up
• Write as Me included

Without Pro, the free tier gives you 10 minutes of transcription every month — enough to keep winWhisper in your daily toolkit.

Anytime, in one click: open the billing portal from the Pricing page (Manage Subscription) or your dashboard. There you can cancel, switch between monthly and yearly, update your card, or download invoices.

Cancellation takes effect at the end of the period you already paid for — you keep Pro until then.

Yes, forever. Credits you purchased never expire and keep working exactly as before — they're spent on transcription when you're not on Pro, and on Live minutes beyond the included 300 when you are.

New credit packs are available to Pro subscribers as real-time top-ups.

Yes, winWhisper requires an internet connection — it connects to cloud-based AI models for fast, accurate transcription, rewriting, and live streaming.

Your desktop app records locally, then sends audio securely to our servers for processing.