SuperWhisper

by Sindre Sorhus

System-wide voice dictation for macOS, Windows, and iOS — hold a hotkey, speak, and AI-formatted text drops into the focused field. Modes rewrite the transcript for email, code, notes, or chat.

Looking at SuperWhisper? Try this first.

Drop your audio. Transcript in seconds. First transcript free, then $2 a file or $8 = 1,000 min

TL;DR

Superwhisper is a native dictation app that wires Whisper (local) and frontier LLMs (cloud, BYOK) to a global hotkey. Hold the key in any app — IDE, mail client, Slack, browser — speak, and the transcript types itself into the focused text field. The differentiator is Modes: each mode is a saved profile of voice model + AI rewrite prompt + auto-activation rule, so the same dictated phrase becomes a tight Slack reply in Slack, a formal paragraph in Gmail, or commented Python in your editor. Built-in modes ship with the app — Voice to Text, Message, Email, Note, Super, Meeting — and custom modes let you write the system prompt yourself.

Best for people who type for a living and want voice as the default input across the OS — engineers dictating commit messages, founders firing off email replies, journalists capturing thoughts into Obsidian. Three tiers: a forever-free plan (small local models, unlimited use), a Pro subscription (cloud + local AI models, BYOK, file transcription, translate-to-English), and an Enterprise plan (SOC 2, centralized billing, model access control). The Free tier also includes a 15-minute trial of Pro features. Verify current pricing on the homepage pricing toggle — monthly / yearly (2 months free) / lifetime are offered; 40% student discount on Pro.

Category
Desktop apps
License
—
Stars
—
Last push
—
Pricing
freemium
Platforms
macOS, Windows, iOS

What it is

SuperWhisper turns Whisper into a system-wide dictation shortcut. Hold a key anywhere on macOS or iOS, speak, and the transcript types itself into whatever text field you're in. Great for email, IDE comments, and chat — not the right tool for transcribing podcasts.

Best for: Replacing Apple dictation with local Whisper — hold a hotkey, speak, text appears.
Watch out for: Not a file-transcription tool; optimized for short dictation, not long-form media.

Install / use

Download SuperWhisper superwhisper.com

Built-in modes · 6 dictation profiles

Each mode bundles a voice model, an AI rewrite prompt, and an auto-activation rule (per-app or per-website). Switch with a hotkey or let the mode trigger itself when you focus a specific app. The list below maps the shipped defaults to the surface they were tuned for — start here, then duplicate one to build a custom mode.

Email mode
Formal correspondence · greeting + sign-off

Rewrites loose speech into a structured email body — paragraph breaks, polite phrasing, kept-on-topic. Auto-activates in Mail, Outlook, Gmail in the browser, or any client you whitelist. Tune the underlying prompt to match your voice (terse vs warm, first-person vs team-voice).


Ships built-in. Pair with a cloud LLM on Pro for tighter rewrites; small local model works on Free.
Message mode
Slack · iMessage · WhatsApp · chat clients

Casual short-form output — keeps contractions, drops filler words, no email-style greeting. Designed for chat surfaces where verbosity feels off. Auto-activate it on Slack and your DM apps and you can dictate replies as fast as you'd type them, without the autocorrect tax.


Light rewrite. Works well with the local small model on Free; cloud models on Pro polish faster.
Note mode
Obsidian · Apple Notes · Bear · journaling

Keeps your raw stream-of-thought largely intact but cleans up false starts and disfluencies. Useful when you want the structure you spoke, not the LLM's interpretation of it. Auto-activate on your notes app of choice for a friction-free capture loop.


Built-in. Disable the rewrite step entirely if you want pure transcription with no model interference.
Super mode + custom code modes
IDE comments · commit messages · git messages

The shipped Super mode handles general-purpose dictation; the recipe for code work is to duplicate it and write a custom system prompt that emits markdown, fenced code blocks, or commented prose. Pair with the Pi coding-agent integration added in v2.13.2 for agent-style follow-on actions.


BYOK on Pro: GPT-5, Claude Haiku 4.5, Llama 4, Grok 4.1, Gemini 3.0 Flash, Ministral, Opus 4.7. Local fallback when offline.
Meeting mode
Long-form capture · transcripts of live audio

The exception to the dictation-only positioning — Meeting mode captures and transcribes long-form audio (your own meetings) and produces a structured transcript. Distinct from the file-transcription path (which accepts MP3/MP4/16 kHz mono WAV from Finder, menu bar, or CLI).


File transcription is gated to Pro; live Meeting capture works on Free with usage limits.
Custom modes
Write your own system prompt

Duplicate any built-in mode and override the AI rewrite prompt, the voice model, the language, and the auto-activation rule (app bundle ID or website URL pattern). This is where Superwhisper diverges from generic dictation — your mode library becomes your personal voice-to-text DSL.


Local Whisper Large for offline; cloud models on Pro for higher rewrite quality.
Notable platform caveat from the vendor: Intel Macs work best with cloud models — offline models only run really well on Apple Silicon. For Whisper-based desktop apps without the AI-rewrite layer see MacWhisper (Mac-only file transcription) and Aiko (free, Mac App Store, drag-and-drop).

Getting started in 3 steps

1Download and grant accessibility access

Pick your platform — macOS 13.3+, Windows 10+, or iOS 18+ via the App Store. The Mac and Windows builds need Accessibility permission so the global hotkey can intercept keystrokes and type the transcript into the focused field.

1. Open https://superwhisper.com/download in your browser.
2. Pick the platform card that matches your OS:
   - macOS: requires 13.3 or later (Apple Silicon strongly preferred for local models).
   - Windows: 10 or later, x64 or ARM64 build.
   - iOS:    requires iOS 18 or later, installed via the App Store.
3. Install via the standard installer flow for your OS.
4. On first launch, the app will prompt for:
   - Accessibility access (so the hotkey can type into other apps).
   - Microphone access (obvious).
5. Grant both, then set your global hotkey in Settings.
Full setup walkthrough: docs · get-started ↗. Hotkey settings: settings-shortcuts ↗.
2Pick your first mode and dictate into a real app

Skip the empty-Notes-window demo. Open the app where you'll actually use voice — your IDE, your mail client, your chat app — and trigger the matching built-in mode. The point of Modes is that the output already fits the surface.

1. Open Mail (or Gmail in the browser) and click into a draft reply.
2. Hold the global hotkey (default varies by OS — set in Settings).
3. Speak: "Sounds good — let's pick this up Thursday afternoon."
4. Release the hotkey.
5. The Email mode rewrites and types the result into the draft.
6. Repeat in Slack with Message mode, in your IDE with a code mode, etc.
7. Set per-app auto-activation under Mode settings so the right
   mode picks itself when you focus that app.
Built-in modes overview: docs · modes ↗. Auto-activation rules let a mode trigger itself by app bundle or website URL pattern.
3Make a custom mode for your highest-volume surface

The single highest-leverage configuration step — duplicate the closest built-in mode and rewrite the system prompt to match your actual writing voice and the formatting your target app expects (markdown, fenced code, bullet lists, specific tone).

1. Settings → Modes → duplicate the closest built-in mode
   (e.g. duplicate Message mode for a custom Slack mode).
2. Edit the AI Prompt field — be specific:
   - Tone you want ("terse engineer voice", "warm but professional").
   - Formatting rules ("emit markdown", "wrap code in fences", "use bullet lists for steps").
   - Constraints ("keep under 3 sentences", "do not add greetings").
3. Pick the voice model — local Whisper for offline / privacy,
   or a cloud LLM via BYOK on Pro for higher rewrite quality.
4. Set the auto-activation rule — match by app bundle ID or by
   website URL pattern so the mode picks itself.
5. Trigger the hotkey while focused on the target app.
Custom modes guide: docs · modes ↗. Bring-your-own-key for cloud LLMs is a Pro feature; supported providers include GPT-5, Claude Haiku 4.5, Llama 4, Grok 4.1, Gemini 3.0 Flash, Ministral, and Opus 4.6 / 4.7 (added in v2.13.2).

What people actually do with SuperWhisper-style transcription

The tool is the means. These are the jobs — each one priced at published rates, each one wired up on its own page.

Transcribe a whole back catalogue in one runThe archive is the asset, but one-at-a-time upload means it never gets done.…1,000 hours ≈ $480, onceTranscribe every new file in a Drive folderFiles accumulate in a folder nobody opens, and the knowledge in them is invisible to search.…≈ $0.48 per audio hourCut short-form clips out of a webinar automaticallyThe clips worth posting are buried somewhere in a 60-minute recording and finding them by scrubbing costs more than the…60-min webinar ≈ $0.48

Wire it into Amazon S3n8nMake.comGoogle DriveZapierSlack

Features

Speaker diarizationNo
Word-level timestampsNo
Streaming / real-timeYes
Languages supported99
HIPAA eligibleNo

Links

SuperWhisper vs Whipscribe

FeatureSuperWhisperWhipscribe
CategoryDesktop appsTranscription APIs
Pricingfreemium$8–$24 one-time packs (credits never expire) · $2 single unlock · free instant preview
Speaker diarizationNoYes
Word timestampsNoYes
StreamingYesNo
Languages9999
PlatformsmacOS, Windows, iOSWeb, API, MCP

Where this category is heading

From the vendor changelogs we track weekly — what changed in August 2026, and what it means if you are choosing now.

Insight on top of the transcript, not sold separately

AssemblyAI moved summarisation onto an LLM this month; every vendor is racing to return action items, quotes and topics with the text rather than as an add-on.

Whipscribe today Every Whipscribe job already returns an insights payload — summary, key quotes, topics and speakers — from the same job id, at no extra charge.

Private, secured deployment

Deepgram shipped self-hosted container images in August — the market is moving toward audio that stays inside a boundary the customer controls, because teams with customer calls or unreleased material are refusing shared model endpoints.

Whipscribe today Whipscribe runs on our own GPUs in a private, secured cloud. Audio is never forwarded to OpenAI or any third-party model.

Agents calling transcription as a tool

The fastest-growing way to use a transcription API is not a form — it is Claude, Cursor or a workflow runner calling it mid-task through MCP.

Whipscribe today Whipscribe ships an MCP server: transcribe, search and summarise from an assistant without wiring anything.

Async and batch over one-file-at-a-time

AssemblyAI's 1.0 SDK unified async, realtime and sync; Deepgram's CLI went to 0.3. The unit of work is becoming the folder or the bucket, not the file.

Whipscribe today Submit with an Idempotency-Key and a batch_id, poll by job, retry safely. The S3 connector runs a whole prefix in one grant.

More languages, unevenly

Deepgram added Afrikaans, Georgian and Armenian and improved a dozen more this month. Coverage is widening while quality still clusters around English and the large European languages.

Whipscribe today 99+ languages auto-detected. Ask for a language explicitly when you know it — auto-detect on a short or noisy clip is the most common cause of a wrong-language transcript.

Source: Deepgram and AssemblyAI changelogs, scanned 2026-08-24.

Alternatives to SuperWhisper

MacWhisper
Jordi Bruin

Polished Mac app for Whisper — the default pick if you're on macOS.

freemium
Aiko
Sindre Sorhus

Free Mac App Store Whisper app — drag, drop, done.

free
Hindenburg PRO
Hindenburg Systems

Broadcast-style DAW built for journalists, podcasters, and audiobook producers.

paid

Frequently asked about SuperWhisper

What is SuperWhisper?

Superwhisper is a native dictation app that wires Whisper (local) and frontier LLMs (cloud, BYOK) to a global hotkey. Hold the key in any app — IDE, mail client, Slack, browser — speak, and the transcript types itself into the focused text field. The differentiator is Modes: each mode is a saved profile of voice model + AI rewrite prompt + auto-activation rule, so the same dictated phrase becomes a tight Slack reply in Slack, a formal paragraph in Gmail, or commented Python in your editor. Built-in modes ship with the app — Voice to Text, Message, Email, Note, Super, Meeting — and custom modes let you write the system prompt yourself.

Is SuperWhisper free?

Partly. SuperWhisper has a free tier with paid plans above it — published pricing: freemium. Free-tier limits change more often than this page does; verify on the vendor's pricing page.

Does SuperWhisper work on Mac, Windows and Linux?

SuperWhisper runs on macOS, Windows and iOS.

How do I get started with SuperWhisper?

There is nothing to install — SuperWhisper is used in the browser. Create an account on the vendor's site and upload or connect your audio.

How many languages does SuperWhisper support?

SuperWhisper lists 99 languages. That is the Whisper-family multilingual set, so quality varies by language — English and the large European languages are strongest.

Does SuperWhisper identify different speakers?

No. SuperWhisper does not label speakers; a conversation comes back as one continuous text. If you need speaker labels, that is a separate tool or a different service.

Does SuperWhisper give word-level timestamps?

No — SuperWhisper produces segment or sentence timing only, so subtitle cues can land mid-word.

Can SuperWhisper transcribe live audio?

Yes — SuperWhisper supports streaming / live input, not just finished files.

Is SuperWhisper HIPAA compliant?

SuperWhisper does not claim HIPAA compliance. For PHI, run it on infrastructure you control or choose a service that will sign a BAA.

What are the limitations of SuperWhisper?

Not a file-transcription tool; optimized for short dictation, not long-form media.

Who is SuperWhisper best for?

Replacing Apple dictation with local Whisper — hold a hotkey, speak, text appears.

What should I know before choosing SuperWhisper?

Best for people who type for a living and want voice as the default input across the OS — engineers dictating commit messages, founders firing off email replies, journalists capturing thoughts into Obsidian. Three tiers: a forever-free plan (small local models, unlimited use), a Pro subscription (cloud + local AI models, BYOK, file transcription, translate-to-English), and an Enterprise plan (SOC 2, centralized billing, model access control). The Free tier also includes a 15-minute trial of Pro features. Verify current pricing on the homepage pricing toggle — monthly / yearly (2 months free) / lifetime are offered; 40% student discount on Pro.

Does SuperWhisper run offline?

SuperWhisper runs on your own machine as a desktop app. Check whether the model runs locally or calls a cloud service — some desktop tools do both.

Can I use SuperWhisper in the browser?

No — SuperWhisper is a macOS, Windows and iOS application, not a web app. You install it rather than sign in to it.

Who makes SuperWhisper?

SuperWhisper is made by Sindre Sorhus.

Whipscribe is a managed faster-whisper + whisperX service. If you want transcripts without running infrastructure, paste a URL or drop a file in the form below — you'll have a transcript in seconds.

Explore

All transcription tools · Audio technology hub · Transcribe any platform · Audio & video formats · How-to guides · Glossary · Playbooks · Apps · Broadcast & radio · Podcast transcripts · Use cases · Blog · Transcription API · Integrations · Automations · For your industry