Hold one key and speak. The words go straight into whatever you are already typing in, in 99 languages, on your device or through your own API key.
3,215
102,971
51h 13m
17h 4m
Groq
FEATURES OF THIS STAGE
Groq
FEATURES OF THIS STAGE
Optional — improves the text but adds processing time.
I need to review the client proposal before tomorrow's 10am meeting.
Let the team know the deploy ships Friday, after the tests.
Call summary: they want the pilot running by the end of the month.
Auto-detected, or pick one
Speaking beats 45 words a minute
Whisper runs on your machine
No subscription, ever
Right Command on Mac, Right Ctrl on Windows. Prefer not to hold? Switch it to tap to start, tap to stop.
Long thoughts, proper names, technical terms, two languages in one sentence. Speakly transcribes what you actually said.
The text appears at your cursor, in the app you are already using. Nothing to copy, nothing to paste.
Add product names, client names, acronyms and technical terms once. Available when you transcribe with your own API key.
Read the docsFind every place we call the old pricing endpoint and list them.
Terse in the terminal, warm in email. Write a prompt per app and Speakly applies the right one automatically.
Read the docsSelect any text in any app and press Option and a digit to polish, shorten or translate it in place.
Read the docshey can you take a look at the migration branch when you get a sec
Could you take a look at the migration branch when you get a second?
Double-tap the key and ask about the email, page or thread in front of you. It reads the window, never a screenshot.
Read the docsYour microphone and the room audio are captured as two separate tracks, so you can tell who said what.
Read the docsDrop in an audio or video file, or paste a YouTube link, and get the full transcript back.
Read the docsNot a setting buried in a menu. It is the choice the whole app is built around.
Whisper runs locally. No key, no account, no network. It keeps working on a plane. It transcribes you faithfully; rewriting needs a model you set up.
Speech to speech on Teams, Zoom or Meet. No captions for anyone to read, in either direction.
Bom dia! Consegui revisar a proposta e tenho alguns ajustes para sugerir.
Good morning! I've reviewed the proposal and I have a few adjustments to suggest.
Requires your own OpenAI key
This feature uses OpenAI's realtime translation model — it's the one feature that doesn't run on the local models or on other providers. You pay OpenAI directly, around $0.04 per minute (a 30-minute call costs about $1.11), and nothing extra to us.
Try everything for free. Pay once — no subscription.
Try everything, risk nothing
Lifetime license, no subscription
You get 7 days of full access to the application. You can use all features, including Local Models and BYOK. After 7 days, the application becomes read-only until you purchase a license.
Download it, hold a key, watch the words land. Free for seven days, then twenty dollars once.
macOS and Windows. 7-day free trial, then a single $20 payment.
