Repository navigation
Releases: theodoremca/telekey
Release list
TeleKey 0.2.3
Install
macOS (Apple Silicon and Intel): open the .dmg and drag TeleKey onto Applications.
This is a staging build: it signs in to TeleKey's test server, which only invited testers can use.
Windows: run the -setup.exe. It is not code-signed yet, so SmartScreen may warn: choose More info, then Run anyway.
Linux (X11, not Wayland; new and little tested): install the .deb with sudo apt install ./TeleKey_*.deb, or make the .AppImage executable and run it.
Full Changelog: v0.2.2...v0.2.3
TeleKey 0.2.2
Install
macOS (Apple Silicon and Intel): open the .dmg and drag TeleKey onto Applications.
This is a staging build: it signs in to TeleKey's test server, which only invited testers can use.
Windows: run the -setup.exe. It is not code-signed yet, so SmartScreen may warn: choose More info, then Run anyway.
Linux (X11, not Wayland; new and little tested): install the .deb with sudo apt install ./TeleKey_*.deb, or make the .AppImage executable and run it.
Full Changelog: v0.2.1...v0.2.2
TeleKey 0.2.1
Install
macOS (Apple Silicon and Intel): open the .dmg and drag TeleKey onto Applications.
This is a staging build: it signs in to TeleKey's test server, which only invited testers can use.
Windows: run the -setup.exe. It is not code-signed yet, so SmartScreen may warn: choose More info, then Run anyway.
Linux (X11, not Wayland; new and little tested): install the .deb with sudo apt install ./TeleKey_*.deb, or make the .AppImage executable and run it.
Full Changelog: v0.2.0...v0.2.1
TeleKey 0.2.0
Install
macOS (Apple Silicon and Intel): open the .dmg and drag TeleKey onto Applications.
Windows: run the -setup.exe. It is not code-signed yet, so SmartScreen may warn: choose More info, then Run anyway.
Linux (X11, not Wayland; new and little tested): install the .deb with sudo apt install ./TeleKey_*.deb, or make the .AppImage executable and run it.
Full Changelog: v0.2.2...v0.2.0
v0.1.1 — usage tracking
Push-to-talk dictation for macOS. Hold a shortcut anywhere, speak, release, and polished text lands at your cursor.
New in this release
- Usage tab — what you have spent and how many minutes you have dictated, by day, month and lifetime, with a 30-day chart. It records the billing units OpenAI reports back rather than estimating locally, and the rates are editable since published prices change.
- Fixed — the shimmer shown while transcribing escaped the edges of the overlay capsule.
Install
Download Flowtype_0.1.1_arm64.dmg, open it, and drag Flowtype onto Applications.
macOS will say "flowtype is damaged and can't be opened".
It is not damaged. This build is signed but not notarised, so macOS quarantines it. Clear the flag:
xattr -dr com.apple.quarantine /Applications/flowtype.appThen open it normally.
Apple Silicon only. This build will not run on an Intel Mac — build from source instead.
You will need
- macOS 12 or later
- An OpenAI API key with credit on the account
Transcription costs roughly $3–6 a month at heavy use, billed to your own OpenAI account. There is no Flowtype account and no server.
Setting up
The Settings window links to each step and shows a status dot for all of them.
- API key — paste it into Settings, where it goes to your Keychain
- Microphone — prompted on your first dictation
- Accessibility — required to paste into other apps; Flowtype prompts on launch
Check everything at once:
/Applications/flowtype.app/Contents/MacOS/flowtype checkWhat it does
Hold ⌃⌥Space, speak, release. A capsule appears with a live waveform of what the microphone is hearing. Roughly 3–4 seconds from release to text, most of it the transcription request.
Transcription is OpenAI's gpt-transcribe, which punctuates, capitalises and drops filler words itself — so there is no cleanup pass in the way.
Optional, and off unless you configure it:
- Vocabulary — names and jargon sent as recognition hints, so they come back spelled right
- Formatting — per-app styles: literal in a terminal, terse in Slack, formal in Mail
- History — recent transcripts, click to copy
Privacy
Audio is captured to memory, uploaded for transcription, and the buffer wiped — it never touches disk. Transcripts and usage counts stay on your Mac, owner-readable only, and both can be cleared or switched off. Your API key lives in the Keychain and is never logged.
Known limitations
- Apple Silicon only
- Not notarised, hence the quarantine step above
- The trigger must be a modifier plus a key — holding
Fnalone is not supported yet - Usage totals start when you install; there is nothing to backfill from
v0.1.0
First release. macOS 12+, Apple Silicon.