The best dictation apps for Mac (2026)

Full disclosure up front: I make one of the apps on this list (Acousmos), so read this the way you'd read any founder's roundup — I've tried to keep every factual claim checkable against each vendor's own site, and I'll tell you plainly which tools fit which people, including when the answer isn't mine. Prices and tiers below are as of July 2026; check the vendor's pricing page before you buy. For scale: my own Stats panel shows 12,804 dictations and 556,227 words over the past three months — this list comes from daily-driver use of the category, not an afternoon of testing.

The quick answer

• You want your dictations to accumulate into something — a searchable, permanent record of what you said: Acousmos
• You want deep configuration and local models with a free tier: Superwhisper
• You want one cross-platform subscription across Mac, Windows, and phone: Wispr Flow or Typeless
• You want to pay once, not subscribe, and keep everything offline: VoiceInk
• You want the lowest published subscription price: Aqua Voice
• You dictate a few short messages a day: the built-in Apple Dictation is free and already on your Mac

Comparison at a glance

Acousmos — if your words should outlive the moment

I built Acousmos around one conviction: what you say is worth keeping. Every dictation — the original audio plus the text — lands in a permanent archive on your Mac, full-text searchable in English and Chinese, in an open format you can export any time. The plan you talked through in March is three seconds away in July, replayable in your own voice.

The dictation itself works the way you'd expect from this category, plus a few things I haven't seen together elsewhere: four clean-up styles (from verbatim to fully shaped) that switch automatically per app, voice translation (set a target language once — speak Chinese, English lands at your cursor), and a dictionary that learns from your corrections. Speech runs through managed cloud by default; Pro can bring its own API keys or run fully local models, so airplane mode works.

Honest limits: Apple Silicon and macOS 14+ only, no mobile app, and it's subscription-priced above most of this list. If you don't care about keeping what you say, cheaper tools below will serve you fine. There's a one-month free trial with no account, and you can test the engines in your browser at acousmos.com/arena.

Superwhisper — configuration depth and local models

Superwhisper runs on macOS, iOS, and Windows, offers local models on a permanent free tier, and supports bringing your own API keys. Per-app modes and a history view are there too. One caveat on cost: as of mid-July 2026 their site no longer publishes pricing at all — you'll see the paid tier's price inside the app. If you enjoy tuning your tools — picking models, shaping modes — it gives you the most knobs of anything here, and the free tier means you can find out at no cost whether the workflow sticks. We keep a factual side-by-side at Acousmos vs Superwhisper.

Wispr Flow — one subscription across Mac, Windows, iPhone

Wispr Flow covers macOS, Windows, and iPhone under one account (free tier 2,000 words/week; Pro $15/mo, $12 on annual). Transcription happens on its cloud servers — there's no offline mode and no bring-your-own-key option, so weigh that against your own privacy requirements. If your writing happens on three devices and you want zero setup, it's a reasonable shortlist entry. Side-by-side: Acousmos vs Wispr Flow.

Typeless — the largest published free allotment (8,000 words/week)

Typeless gives you 8,000 words a week free, apps on all four major platforms, and a 30-day Pro trial without a card. Pro is $12/mo billed annually — the monthly plan is $30, so the annual commitment is where its pricing makes sense. Like Wispr Flow it processes speech in the cloud, with no offline mode and no BYOK; it does keep an on-device history. Chinese and mixed Chinese-English input are supported. Side-by-side: Acousmos vs Typeless.

VoiceInk — pay once, own it, stay offline

VoiceInk is a one-time purchase ($25–49 depending on tier) and open source (GPL-3.0). Transcription runs entirely on local models, and it keeps a searchable local history with audio. AI text enhancement is optional and runs through your own API keys, which means extra setup and per-use cost — the trade for not paying a subscription. Apple Silicon with macOS 14.4+ only. If subscriptions annoy you on principle and you're comfortable with a bit of configuration, this is the one to try. Side-by-side: Acousmos vs VoiceInk.

Apple Dictation — the free baseline

Before paying for anything, know what your Mac already does: system dictation is free, runs on-device on Apple Silicon, and works offline. What it doesn't do: no AI clean-up or formatting (you get your words literally, filler and all), no history of any kind, no custom vocabulary, no per-app behavior. For a few short messages a day it's genuinely enough — the paid category exists for people who write a lot and want the editing pass to disappear. More detail: Acousmos vs Apple Dictation.

Aqua Voice — the low-cost cloud option

Aqua Voice starts at $8/month billed annually ($96/year; a higher tier runs $12/month annual), covers Mac, Windows, and iPhone, and processes speech in the cloud. Their site lists a transcript-history feature (I have not tested it). I've spent the least time with this one, and its site publishes little technical detail — nothing about offline modes or bringing your own keys — so if those matter to you, verify before subscribing. At the lowest published subscription price on this list, it's the budget cloud pick. Side-by-side: Acousmos vs Aqua Voice.

How to actually choose

Three questions sort this list faster than any feature grid:

1. Do you want to keep what you say? If yes, the field narrows to the tools with real archives — Acousmos (audio + text, permanent, searchable) or VoiceInk (local history). If no, skip to price and platforms.
2. Where should your speech be processed? Fully-local-always: VoiceInk or Superwhisper's local models. Flexible per situation (cloud for accuracy, local when it matters): Acousmos Pro or Superwhisper. Don't care: any cloud tool.
3. One device or many? Mac-only tools (Acousmos, VoiceInk) go deeper on the Mac; Wispr Flow and Typeless spread wider across devices.

And whatever you pick: dictate for two real workdays before deciding. Every tool here demos well; the differences show up on day two, in your actual apps, with your actual vocabulary.