Roadmap

Where Lispr is going.

Lispr is small on purpose — but that doesn't mean it stands still. Here's what's shipped, what we're building right now, and what comes next. No fixed dates — we publish a date once we trust it.

Shipped

Available now

What you can install today.

Lispr for macOS

Live · 71 releases in

The original — where Lispr started.

macOS 11 Big Sur or newer, Apple Silicon & Intel — ~6 MB download, menu-bar only
~99 languages auto-detected, switching mid-sentence
Push-to-talk on right ⌥ by default, any modifier or Fn if you prefer; double-tap latches hands-free; Esc cancels
~0.35 s median end-to-end, measured on the client since streaming uploads landed in 0.41
Smart sentence joining, clipboard-safe, Sparkle auto-updates

Lispr for Windows

Live · early access

Same idea, same hotkey-and-go feel — now shipping. Download for Windows.

C# / .NET 8 with WPF — small native bundle, no Chromium tax
Azure Trusted Signing for a clean SmartScreen install
Velopack for delta auto-updates (the Sparkle equivalent on Windows)
Same Whisper-on-edge transcription path — no sign-up, same as on macOS

Up next

Designed, queued

Architecture decided, plan written, build hasn't started.

Lispr for iPhone & iPad

Designed

iOS has no global hotkey — so Lispr lives as a system keyboard you switch to. Universal app, both phone and iPad on day one.

Custom system keyboard with a big mic button — works in any app that takes text input
Container app for permissions, history, and your usage stats
Free, App Store-distributed, no sign-up

Lispr for Android

Designed

Android does support a global trigger — via an Accessibility Service and an overlay bubble, like Wispr Flow ships now.

Floating mic bubble you can move anywhere on screen
Accessibility Service inserts text into whichever app you're focused on
Play Store-only first; sideloading later

Also exploring

Investigating, not committed

Things we'd like to do if they make sense. Tell us if any of these matter to you.

Wake word. Say "Lispr, …" to start a dictation without touching a key. In training on shared fragments; nothing is promised yet.
Voice commands beyond dictation. "New line", "delete that", "open Slack" — short imperatives that go through a command parser instead of the transcription path.
Public API. The Lispr edge proxy as a service for other developers who want sub-300 ms transcription with our latency and privacy posture, without running their own Groq integration.
Team dictation. Shared custom vocabularies and usage reporting for design / dev / legal teams that all use Lispr.

Not on the roadmap

Deliberately not building

These cut against what Lispr is. Saying so explicitly so no one waits for them.

Always-listening mode. Push-to-talk is the model. Listening in the background is a different product with different privacy trade-offs — not ours.
Cloud storage of your transcripts. Your dictations live where you typed them. The server forgets the audio the second the text comes back; recent recordings stay on your own Mac so nothing is lost if the app dies mid-dictation. We're not adding a "history in the cloud" feature.
Mandatory accounts. Dictation never requires signing in. The optional account exists for one thing — carrying a paid plan across your devices — and even that is a one-time email code: no password, no profile.
Ads, or profiling you. No ad networks, no third-party trackers, and nothing you dictate is used to build a profile of you. Lispr does send its own diagnostic events — timings, error codes, which onboarding step you reached — so we can see what breaks; there's a switch in the menu to turn that off.

Want something here?

We genuinely read every request. If something in Also exploring would change your day, or if there's something missing from this page that we should think about, tell us.

[email protected]