Talk. It types itself. In any language.
Hold ⌥ to dictate. Add ⌃ to translate. Lands at your cursor in any Mac app.
Download — it's freefree · macOS 11+ · no sign-up
Notarized by Apple · made by Codebridge
One gesture. Two languages.
Hold a key to talk in yours. Tap a second key mid-speech to land in theirs.
Hold to talk
Press and hold the right Option key. Speak in your native language — Spanish, Portuguese, Japanese, Hindi, whatever.
Tap to translate
Add a second key mid-speech to pick your target language. Skip it if you just want dictation.
It lands typed
Translated text appears right at your cursor. No copy-paste, no app switch, no chat UI in the way.
The list is real, not marketing.
Your speech language is detected automatically — around 99 of them. These 32 are what Lispr translates into.
Stop talking. Under a second later, the translation is already typed.
Said in English. Sent in Spanish.
Hold ⌥+⌃, speak English, send Spanish.
Does the boring stuff well too.
Plain dictation too
Skip the second key — same gesture works for monolingual dictation. Same speed, same languages, your cursor.
Custom vocabulary
Add product names, code identifiers, your client list. Both sides — dictation and translation — respect the dictionary.
Instant
~0.35s for dictation, ~0.7s for translate. No "thinking…" spinner — text just shows up.
Any app
Slack, Notes, Figma, Terminal. If a cursor blinks, Lispr types there.
Clipboard-safe
Inserts your text, then restores what you'd copied.
Lives quietly
Menu bar, push-to-talk, one key. No window, no popups, no clutter.
Where Lispr fits.
| Lispr | Wispr Flow | Apple Dictation | |
|---|---|---|---|
| Price | Free · Pro $7.99/mo | $15/mo | Free |
| Translation built-in | ✓ | — | — |
| Works without an account | ✓ | — | ✓ |
| Languages, auto-detected | ~99 | ~100 | manual |
| Push-to-talk, one key | ✓ | ✓ | — |
| Footprint | ~17 MB, menu bar | full app | built-in |
| Speech model | Whisper large-v3-turbo | proprietary | on-device |
Public details as of August 2026. Lispr does one job, fast.
Safe to install, private by default.
Checked by Apple
Every release is notarized by Apple — scanned for malware, so macOS opens Lispr without a security warning.
A real, named company
Built by Codebridge — a real software company with a real address, not an anonymous developer. Every release is signed with an Apple Developer ID and notarized by Apple before we publish it.
No sign-up required
Lispr works out of the box — no account needed. The optional account (it carries a paid plan across your devices) is just an email code: no password to steal.
Your voice is transcribed, then gone
Audio travels over an encrypted connection, gets transcribed, then we discard it. We keep no recording and none of it trains a model. The transcription provider holds it for up to 30 days for abuse review before deleting it. Sharing recordings is a setting you turn on yourself; it's off unless you do.
Translation included
It's part of Lispr, not a paid add-on — translation is in the free plan, and that's a commitment, not a launch promo. Pro removes the weekly minutes cap; it never locks away languages or translation.
What people say once they've used it.
“This thing is really awesome. I started using voice dictation for notes and texts so I don’t have to type on a keyboard. The voice recognition is just amazing and fast, which saves a ton of time.”
“The ‘No-Account’ policy is a breath of fresh air. In an era where every utility demands your email and marketing consent, Lispr respects the developer workflow: install, run, and it just works.…”
“The recognition is impressively accurate, even when I’m speaking naturally instead of dictating word by word. It feels much faster than typing, especially when I already know what I want to say.”
The honest answers.
What permissions does Lispr need?
Two, each granted once in System Settings: Microphone — so it can hear you — and Accessibility — so it can type into your other apps. Lispr walks you through both on first launch.
Which Macs are supported?
macOS 11 Big Sur or newer, on Apple Silicon and Intel. The download is about 6 MB; installed it takes around 17 MB.
Does it work offline?
No — transcription runs in the cloud, which is why it's this fast and accurate. From letting go of the key to text at your cursor is about 0.35 s.
Which languages does it understand?
Around 99 for understanding what you said — that's Whisper, our speech model. For translating into another language, we polish 32 of those (see below).
Is it really free?
Yes — dictation is free, and stays free. The free plan comes with a weekly allowance of dictation minutes that covers most people comfortably; if you dictate all day, Lispr Pro ($7.99/month, or $79/year) removes the cap. Translation, custom vocabulary and all languages stay in the free plan. And the first 1,000 early active users keep Pro free forever — see Founding.
How is translation different from dictation?
Same gesture, one extra key. Hold ⌥ to dictate — Lispr types what you said verbatim. Add ⌃ to translate — Lispr types your sentence in your target language instead. You set your top two target languages in Settings, one per slot key.
Which target languages can Lispr translate to?
32 — the languages most people on Earth actually speak. Pin your top two to slot keys; pick any of the rest in Settings.
Stop typing in a language you don't think in.
Hold a key. Speak yours. Send theirs.
Download Lispr — freemacOS 11+ · ~6 MB download · no sign-up