Dictate Keyboard (Whisper AI)

DevEmperor

About this app

Dictate is a simple yet powerful keyboard, which converts spoken language into text. It uses an extremely powerful speech recognition AI in the background and therefore supports all common national languages. What's more, you don't have to dictate punctuation because Whisper adds it itself. You can even sing in text and it will be recognized correctly.

Other features include:
- custom rewording instructions (translation, summarizing, style-changing, ...) using GPT-5 (and many other models) 💡
- auto formatting of spoken commands ("new paragraph", "first, second, ...", "Henry with i", etc.) ✨
- switching between different transcriptions models (Whisper, GPT-4o-transcribe, ...) and providers (OpenAI, Groq or your own transcription server) 🔄
- transcription of voice and audio or video files 📄
- recording using phone or Bluetooth microphone 🎧
- easy emoji and number insertion 🥳
- export and import of all your prompts 📤
- setting a system prompt to guide the style of the speech recognition and rewording ✏️
- dark-mode, light-mode and custom theme settings ✨
- direct input, without the typing animation ⏱️
- quick switching to the standard keyboard ⌨️
- haptic feedback 📳
- custom API server and proxy support 🛠️
- simple use and setup process 💪

For the app to be fully functional, every user must create an API key with OpenAI or Groq when setting it up for the first time. This incurs very low fees for each transcription (usually a few cents per month). There is a detailed explanation of this during setup.

Dictate is completely open source, the source code is available on GitHub.
https://github.com/DevEmperor/Dictate

If you have any bugs or suggestions for improvement, just write me an e-mail or open an issue on GitHub.

If you like Dictate, I would really appreciate a donation via PayPal. You can find my PayPal-Link on my GitHub page. 😇

Have fun with Dictate! 😊

Ratings and reviews

This thing is like having ChatGPT Whisper 3 with no API cost. You're amazing. I have a form of Dyslexia where I can't spell, but I'm using it and it's amazing. It's worth the four dollars. Buy it. The only downside is it not having a full keyboard. If you cannot read a word and you need to spell it, and you cannot pronounce it right, that's the only downside.

As someone who uses voice to text a lot, I just wanted to say this is really a great thing. I've found myself not having to go back and actually correct mistakes or add punctuation. I don't have to chop my sentences up by actually saying comma or period or exclamation mark. All of these things are really great. I think this tool is amazing. Good job. P.S. I wrote this in one shot and I don't see a single mistake using this keyboard.

I am a horrible speller and am terrible at using proper grammar. So I use voice to text a lot. When I got a OnePlus 13R, the voice to text was horrendous. No matter what keyboard I downloaded, I decided to give this a try even though it wasn't free. So far the difference is night and day. I've used it for a couple days now and I am so impressed and wish I would have gotten it sooner.

What’s new

v3.2.0

- You can now save separate API keys for each provider (OpenAI, Groq, Custom), making switching between them seamless. 🔑
- Connecting to Bluetooth headsets is now faster and more reliable. 🎧
- Added an option to automatically press Enter/Send after transcription. ⚡
- Added support for OpenAI o1, o4-mini, and GPT-5.2, including updated pricing info. 🤖
- Various bug fixes and improvements. ✨

I wish you a blessed New Year 2026! 🎆