The best speech-to-text apps in 2026
Nine ways to turn speech into text, A to Z — and they are not competing with each other, because half of them are for meetings and half are for typing.
The single most useful question in this category is where the words end up. A meeting transcriber puts them in a document you go and read; a dictation tool puts them at the cursor you were already in. Those are different products that share a model, and buying the wrong one is the most common mistake here.
The second question is what happens between the microphone and the text: whether filler words are removed, whether punctuation comes from context, whether the language can change mid-sentence, and whether you can watch the words arrive or only wait for them. Listed alphabetically.
The list below is in alphabetical order, not ranked. aiCat is one of the tools in it, so any ranking we published would be worth what you would expect — instead, every entry says plainly which situation it is the right answer for, and which it is not.
What was actually compared
- Where the transcript lands — at your cursor, or in a document
- Whether the words appear as you speak or only when you stop
- Which applications it covers, and on which operating systems
- Languages, and switching between them inside one recording
- What an hour of dictation actually costs
All 9, A to Z
| Tool | Price from | What it is | Languages | Strongest at |
|---|---|---|---|---|
| aiCat | Free · Premium from $6/mo | Every app, OS-wide | Any language the model speaks | One key that proof-reads, translates, explains, detects, dictates and reads aloud |
| Apple Dictation | Free | Built-in dictation | Several dozen languages | Free, already installed, and works offline on recent hardware; Types into any macOS text field |
| Dragon | One-off from ~$500 · mobile from ~$15/mo | Professional dictation | Several, English strongest | Accuracy after training on your voice, still the benchmark in medical and legal dictation; Deep voice control of the computer, not only text entry |
| MacWhisper | Free · Pro one-off from ~$59 | Transcription app | Multilingual (Whisper models) | Whisper running locally on your Mac, so audio never leaves it; One-off price instead of a subscription |
| Otter.ai | Free (capped) · Pro from ~$8.33/mo | Meeting transcription | English-first | Meeting transcription with speakers separated and searchable; Joins Zoom, Meet and Teams by itself |
| Superwhisper | Free · Pro from ~$8.49/mo | Dictation app | Multilingual (Whisper models) | Local, offline dictation that types into any macOS application; Modes that post-process what you said for a chosen context |
| Whisper (OpenAI) | Free to self-host · API per minute | Speech model | 90+ languages | Very good multilingual transcription for the price; Open weights, so it can run entirely on your own hardware |
| Windows Voice Access | Free | Built-in dictation | Several, English strongest | Free and built in, with full voice control of the OS as well as text; On-device processing on recent builds |
| Wispr Flow | Free (word-capped) · Pro from ~$12/mo | Dictation app | Many, English-first | Dictation typed straight into whatever application you are in; Tidying the filler words and false starts out of what you said |
The list, A to Z
aiCat
For dictation in more than one language, watched as it lands
Why: Each phrase is typed at your caret as you say it, in any application on macOS, Windows and Linux, and one recording can change language halfway. The same keys then proof-read or translate what you just said. Free to start, Premium from $6/month.
Watch out: English cleanup is less aggressive than the specialists', and it is a download with a microphone permission.
Download for free →Apple Dictation
The free baseline on a Mac
Why: Already installed, works offline on recent hardware, types into any text field, and there is no account and no subscription to think about.
Watch out: You say 'comma' out loud, nothing removes filler words, and the language is whatever system settings say.
How aiCat's dictation works →Dragon
For dictation that is part of a regulated job
Why: After voice training it remains the accuracy benchmark for medical and legal work, with custom vocabularies and full voice control.
Watch out: A one-off cost in the hundreds, complicated licensing, an afternoon of setup, and a withered Mac story.
How aiCat's dictation works →MacWhisper
For turning recordings into documents, locally
Why: Batch transcription with speaker detection and subtitle export, running Whisper on your own Mac so the audio never leaves it. One-off Pro licence from about $59.
Watch out: macOS only, and it is a transcriber of files rather than a dictation key.
How aiCat's dictation works →Otter.ai
For meetings, which is a different product entirely
Why: Joins Zoom, Meet and Teams by itself and produces a searchable, shareable transcript with summaries and action items.
Watch out: It does not type at your cursor and is not a dictation tool. English-first, with minute caps below Pro.
How aiCat's dictation works →Superwhisper
For dictation that never touches a network
Why: Local models typing into any macOS application, with post-processing modes per context and a lifetime licence available.
Watch out: macOS and iOS only, costs memory and battery, and nothing happens to the text afterwards.
How aiCat's dictation works →Whisper (OpenAI)
The model most of this list is built on
Why: Open weights, 90-plus languages, free to run yourself, and billed per minute if you use the hosted API — the cheapest transcription here by a distance.
Watch out: There is no application: no hotkey, no caret, no punctuation from context. You are building the product.
How aiCat's dictation works →Windows Voice Access
The free baseline on Windows 11
Why: On-device on recent builds, types into any application, and controls the OS by voice for anybody who needs that.
Watch out: Windows 11 only, behind the paid apps on quality, and no cleanup of what you said.
How aiCat's dictation works →Wispr Flow
For English dictation that reads as though you wrote it
Why: The cleanest output on this page: filler and false starts removed as standard, a personal dictionary for names, and an iPhone keyboard nothing else here has.
Watch out: English-first, no Linux, a waveform instead of words while you talk, and a free tier metered by words per week.
aiCat vs Wispr Flow →Hands-on notes
First-person impressions from using these, not a benchmark. They are here because everything above this point is a table, and a table cannot tell you which tool you will still be opening in three months.
I dictate a lot and in two languages, which rules out most of this list immediately. Setup took a few minutes and asked for the microphone once. The thing that turned out to matter more than accuracy was watching the words land phrase by phrase — a misheard name gets caught in the second sentence instead of the tenth. The other half is that dictation is not a separate app here: the same keys proof-read and translate what I just said, and it remembers the names and terms I have used before and gets them right. It is a download, with an installer and a permission — which is what makes it work at every cursor rather than in one window.
On English it produces the cleanest prose of anything I tried, and there were paragraphs I sent untouched. On Russian it was not usable for me, and on the recordings that switch language mid-sentence — which is how I actually talk — it had nothing to work with.
Bought for the wrong reason and kept for the right one. It is useless as a dictation tool and excellent at what it is actually for: a searchable record of a call that somebody who missed it can read.
Free, instant, already installed, and the honest baseline everything else has to beat. It beats it by having to say 'comma' out loud.
One of them is a hotkey. Try it on the text you have open
Free to start with 10 aiCatCoins, no card. Premium starts from $6/month. macOS, Windows and Linux.
aiCat pricing at a glance
| Plan | Price | aiCatCoins / month | Max selection | What it unlocks |
|---|---|---|---|---|
| Free | $0 | 10 | 500 chars | Everything aiCat does, plus free dictation and your own memory |
| Premium | from $6/mo (list $10) | 3 000 | 5 000 chars | Think More, read-aloud, encrypted memory, premium models |
| Team + Offline | from $18/member | 12 000 | 50 000 chars | Edit your memory, MCP access, local models, team sync |
| Enterprise | from $90/member | 150 000 | Unlimited | SSO, audit, best coin rate |
Shortlist of one? Start with the one that costs nothing to try
aiCat's free tier includes every task, 10 aiCatCoins a month, and coins bought outright never expire. macOS, Windows and Linux.
Frequently asked
What is the difference between dictation and transcription?
Dictation types into the field your cursor is in, live. Transcription turns a recording or a meeting into a document afterwards. Otter is the second kind; most of the rest are the first.
What is the cheapest good dictation?
Whatever is already on your computer — Apple Dictation or Windows Voice Access — costs nothing and is competent for plain text. Paying buys cleanup, punctuation from context, and multilingual handling.
Which handles more than one language in a single recording?
Very few. Most tools ask you to select a language before you start. Model-based tools that transcribe without a fixed language setting cope with a sentence that switches, which is common in bilingual workplaces.
Which operating systems does aiCat support?
macOS (Apple Silicon and Intel), Windows 10/11, and Linux as AppImage, .deb or .rpm.
Is my text sent anywhere?
Only the text you explicitly act on is sent to the model you selected, for that one request. Your history is stored locally on your device. Team + Offline can run a local model so nothing leaves the machine at all.