Buying Guide
What Is Typeless? Pricing, Privacy and Wispr Flow
Typeless is an AI voice keyboard for Mac, Windows, iPhone and Android. What it costs, what it keeps about you, and how it compares with Wispr Flow.
Typeless is an AI voice keyboard. You press a key or tap the mic, talk, and it types a cleaned-up version of what you said into whatever app you are in, with the ums, repeats and false starts taken out. It runs on Mac, Windows, Linux, iPhone and Android, the free plan allows 8,000 words a week, and Pro is $12 a month billed yearly or $30 month to month.
It is new. The iPhone app reached the App Store in December 2025, and US search interest in the name has grown more than tenfold over the past year, which is why so many people are meeting it for the first time. It is made by Simply CA LLC, a Palo Alto company founded by Stanford alumni and led by Huang Song.
Which Typeless is this?
The name is crowded, so it is worth checking you are reading about the right product. On the App Store the dictation app is listed as Typeless: AI Voice Keyboard. There is also an unrelated text expander called Typeless: Keyboard Shortcuts, and a different AI keyboard called Typeless Flow from a developer in Shenzhen.
If a review talks about snippets and saved templates, it is about the text expander. This article is about the dictation app at typeless.com.
What it actually does
Typeless is not a notes app and not a meeting recorder. It sits in the slot your keyboard occupies: on a Mac it is a hotkey, on an iPhone it is a keyboard you switch to. Whatever you say arrives as text in the app that had focus, whether that is Gmail, Slack, a ChatGPT prompt or a form on a website. It works the same way as Wispr Flow, which is the product most people compare it with.
What it does to your words on the way through is the part that makes it more than built-in dictation:
- It edits as you talk. Filler words and repeated words come out, and when you correct yourself mid-sentence (“Tuesday, no, Wednesday”) it keeps only the version you meant.
- It formats by shape. A spoken list comes out as a list, and the tone shifts with the app, so an email reads more formally than a Slack reply.
- It learns your words. Names, brands and jargon are picked up as you use it, and you can add terms by hand.
- It translates while you speak. Set an output language and dictate in another one.
- It takes commands. Select text and say make this more formal or summarize this in two sentences, or ask a quick question without leaving the app. Some of these commands are desktop only.
It detects more than 100 languages automatically and copes with switching language mid-sentence, which is a point Typeless leans on hard for Japanese, Korean and Chinese speakers.
What it costs
| Plan | Monthly | Billed yearly | What you get |
|---|---|---|---|
| Free | $0 | $0 | 8,000 words a week, standard accuracy, standard access at busy times |
| Pro | $30 | $12 a month ($144 a year) | Unlimited words, enhanced accuracy, priority access at busy times |
| Enterprise | Custom | Custom | Adds SSO, SCIM provisioning, audit logs and enforced HIPAA controls |
The free plan is the generous part. At roughly 250 words to a medium email, 8,000 words a week is about 30 emails, which is enough to use it for real work rather than just to try it. That is four times Wispr Flow’s desktop allowance and eight times its phone allowance.
The monthly price is the part to notice. Billed yearly, Pro costs exactly what Wispr Flow Pro costs. Billed monthly, it costs twice as much.
So the free plan is the right way to decide, and if you decide to pay, the yearly plan is the one that makes sense. Month to month only pays off for a short burst of heavy use.
What it keeps about you
Privacy is where Typeless is most specific, and its data controls page is unusually plain about it:
- Transcription happens in the cloud. Your audio goes to Typeless’s servers for accuracy and speed, so it needs a connection and does not work offline.
- Audio and context are discarded. Along with your voice it reads limited context, such as which app you are in and relevant text inside it, so it can get tone and names right. Both are processed in real time and dropped once the text comes back.
- Nothing is used for training. Dictation data is not used to train models, by Typeless or by the AI providers it uses, which include OpenAI under zero-retention agreements.
- History stays on your device. Your dictation history lives on the device unless you switch on cloud sync, and switching sync off deletes the cloud copy. Typeless keeps anonymous usage figures such as total word count.
The one thing worth knowing that the headline does not say: because it reads context from the app you are in, what is on screen near the cursor travels with your voice for that moment. It is not stored, but it is sent.
Typeless vs Wispr Flow
The core job is the same: both clean up speech, punctuate it and adapt the tone to the app. The real differences are in the terms around the product.
| Typeless | Wispr Flow | |
|---|---|---|
| Free plan | 8,000 words a week on every device | 2,000 a week on desktop, 1,000 on mobile |
| Pro, billed monthly | $30 | $15 |
| Pro, billed yearly | $12 a month | $12 a month |
| Platforms | Mac, Windows, Linux, iPhone, Android | Mac, Windows, iPhone, Android |
| Where it runs | Cloud, no offline mode | Cloud, no offline mode |
| Model training | Dictation data never used | Opt out in settings; locked off on Enterprise |
| Beyond dictation | Translation while speaking, voice commands | A meeting notetaker on Mac and Windows |
| App Store rating | 4.5 from 710 ratings | 4.8 from about 16,000 ratings |
Prices in US dollars and ratings as of September 24, 2026.
Read that table as three trade-offs. Typeless has the bigger free plan and the stricter privacy default: nothing to switch off. Wispr Flow has the cheaper monthly price, a far larger base of users behind its rating, and a $280 million funding round in August 2026 behind the company. If the thing that matters most to you is dictating without a connection, neither fits, and the Wispr Flow vs Superwhisper comparison covers the local option. For the full field on a phone, see the best dictation apps for iPhone.
What a keyboard that keeps nothing cannot do
Zero retention is the right promise for a keyboard. You dictate a salary figure, a symptom, a line about a colleague, and you want the tool that carried it to forget it immediately. Typeless is built to do exactly that.
It also means the tool remembers nothing on your behalf, and that shows up in one specific place: the things people tell you. You come out of a coffee and dictate Maya is moving to Denver in March, and her daughter starts at CU in the fall into whatever app is open. Typeless turns it into a clean sentence and forgets it, as promised. The sentence now lives in Apple Notes, or a text to yourself, filed by date and by app, not by Maya.
That is not a flaw in Typeless. Deciding where your words belong is not a keyboard’s job, and one that tried would be a worse keyboard. It is the reason the second half of the habit needs its own home: something that files what you said under the person it was about, so it is there in March when Maya’s name comes up. That is what voice notes for relationship memory are for, and a fast voice keyboard makes them easier to write than ever.
Key takeaway: Typeless is an AI voice keyboard that types cleaned-up speech into any app on Mac, Windows, Linux, iPhone and Android. The free plan allows 8,000 words a week, Pro is $12 a month yearly or $30 monthly, and it processes in the cloud while keeping no audio and never training on your dictation. Against Wispr Flow it wins on free allowance and privacy defaults and loses on monthly price and track record.
FAQ
Is Typeless free?
Yes. The free plan allows 8,000 words a week on every device, with standard accuracy. Pro removes the cap and adds enhanced accuracy and priority access when the service is busy.
How much does Typeless cost?
Pro is $12 a month billed yearly, which is $144 a year, or $30 a month billed monthly. Enterprise pricing is custom and adds SSO, provisioning, audit logs and enforced HIPAA controls.
Is Typeless safe to use?
Typeless says your audio is not stored, the context it reads is discarded after each dictation, and your data is never used to train models, including by the AI providers it uses such as OpenAI. History stays on your device unless you turn on cloud sync. Transcription does happen in the cloud, so your voice does leave the device briefly.
Does Typeless work offline?
No. Dictation, translation and voice commands all need an internet connection, because transcription runs on Typeless’s servers.
Who makes Typeless?
Simply CA LLC, based in Palo Alto, California. The company says it was founded by Stanford alumni, and its founder and CEO is Huang Song. The iPhone app launched in December 2025.
Is Typeless better than Wispr Flow?
It depends on what you weigh. Typeless has a much larger free plan, never trains on your dictation, and runs on Linux. Wispr Flow is half the price month to month, has a far larger base of users, and adds a meeting notetaker. Billed yearly, they cost the same.
Intriq is where the sentence you just dictated can go: spoken notes file to the person they are about, not to whichever app had focus. Start with the AI relationship assistant hub, or read how voice notes become relationship memory.