Superwhisper
Dictate into an existing editor, choose transcription and cleanup separately, and keep a route back to the original recording.
Sources checked 3 October 2026.
PlansFree; from $8.49/mo Website prices; app-store prices may differ
- Free
- Free
- Unlimited local Whisper dictation, custom vocabulary and text replacements; up to two modes without AI cleanup. New users get a 3,000-word Pro-feature trial.
- Pro
- $8.49/month
- Website price; one license works across supported devices.
- Pro annual
- $84.99/year
- Website price; same Pro features as monthly billing.
- Pro lifetime
- $249.99 one time
- Website price; same Pro feature set.
Pro is $8.49/month, $84.99/year, or $249.99 one-time on the website.
Free keeps unlimited local Whisper dictation; new users receive a 3,000-word Pro-feature trial.
View current plans (opens in a new tab)AI & setupIncluded or BYO KeyInstall and grant microphone access
Download local models before offline use; local language processing depends on compatible device hardware.
Where it worksmacOS, Windows, iOS, AndroidMac, Windows, iOS, and Android
One Pro license covers supported devices; custom modes are saved per device by default.
Usage & termsLocal and cloud processingFull local setup takes both steps
A cloud language model receives the recognized text even when speech recognition runs locally.
File transcription, AI modes, context awareness, and provider API keys require Pro; local model availability differs by device.
Model choices
Choose the transcription and cleanup separately.
A speech model turns audio into text; an optional language model processes that transcript. Local and cloud choices can be combined, or the second step can be skipped for transcription without an AI rewrite.
A fully local setup can keep both stages on the computer. Downloaded speech models continue working offline, while local language processing needs compatible hardware. These choices matter when preserving an unpublished draft’s wording and deciding where its text is processed.
Task-specific modes
Build modes around the writing task.
Modes save model choices, language, and processing preferences. Voice to Text preserves transcription; Message, Mail, and Note add cleanup, while a Custom mode applies the author’s own recurring instructions.
Modes can activate automatically in selected apps or websites. The writer can therefore keep different processing for manuscript dictation, notes, and correspondence rather than changing the settings each time.
Recording history
Return to the original recording.
History retains audio, transcription, and the AI result. An earlier dictation can be reprocessed with a different mode without saying it again, and searchable transcripts make it easier to recover a spoken idea.
The distinction between the original transcript and processed result gives the writer a way back when cleanup changes a line’s wording. Recording retention can be shortened from its default of keeping files indefinitely.
Context, files, and devices
-
Use context deliberately.
-
Pro modes can use the active application, selected text, and recently copied text. Super Mode uses that context for dictation and formatting; Custom modes let the writer specify the task and which context to include.
-
Teach it the vocabulary of the book.
-
Custom vocabulary and text replacements help with names and terms that general speech recognition can mishear. Both are available on the Free plan.
-
Bring recorded research into text.
-
Pro transcribes audio and video files through the active mode, so interviews and voice memos can use the same chosen transcription and formatting. System-audio recording and speaker separation extend the recording workflow.
-
Keep one license across writing devices.
-
Mac, Windows, iPhone, and Android share a Pro license. The free plan retains unlimited local Whisper dictation after the introductory Pro allowance, making ongoing transcription available without a subscription.
Feature details & official sources
Purpose and capability information checked 3 October 2026.
Platform and pricing reference: 24 October 2026.
Recorded requirements and limits
- Install the app, allow the required microphone and accessibility access, and download local models before using them offline. Local model availability and performance depend on platform and hardware.
- Fully local processing needs local speech recognition and local language processing, or no language processing. A local speech model followed by a cloud language model still sends text to the cloud.
- File transcription, AI modes, cloud models, context awareness, and provider API keys require Pro. Speaker-separated transcripts do not automatically pass through the same mode’s AI processing.
- Website prices differ from in-app store prices. One Pro license covers Mac, Windows, iPhone, and Android; custom modes are stored per device by default.
- Current dictation product (opens in a new tab)
- Separate speech and language model choices (opens in a new tab)
- Free and Pro features, trial, prices, and licensing (opens in a new tab)
- Offline processing and browser-demo boundary (opens in a new tab)
- Platforms and minimum operating systems (opens in a new tab)
- Mode settings and automatic activation (opens in a new tab)
- Built-in cleanup and context modes (opens in a new tab)
- Custom instructions and chosen context (opens in a new tab)
- Audio, transcript, and AI result history (opens in a new tab)
- Audio and video file transcription (opens in a new tab)
- Custom vocabulary and replacements (opens in a new tab)
Information may be inaccurate. Confirm details on the official website. Report errors to [email protected].