Dictate 4.0.0

Dictate 4.0.0 is a complete rebuild — Dictate is now a full keyboard! 🎉

• ⌨️ Real keyboard: Dictate is now built on top of a full keyboard (FlorisBoard). No more app switching or accessibility setup — dictate directly into any app.
• 🌐 Many new providers: OpenAI, Groq, OpenRouter, Mistral, xAI (Grok), DeepSeek, Together AI, DeepInfra, Ollama (local) and any custom OpenAI-compatible server — each with its own API key. 🔑
• 🔄 Always-current model list: model catalogs load live from your provider, so new models show up automatically.
• ✨ AI rewording & custom vocabulary: reshape your text with AI and teach Dictate your own words.
• 🗣️ Language while recording: pick and see your input language right on the recording bar.
• 📂 File transcription: transcribe audio files in more formats.
• ⚠️ Clearer error messages: detailed, provider-specific errors when something goes wrong.

Thank you to everyone who reported issues and shared feedback — this release closes a lot of long-standing requests. 💙

Dictate is distributed via the Google Play Store.

Dictate 3.2.0

This update brings some major improvements and new features! 🚀

• Separate API keys: You can now save separate API keys for each provider (OpenAI, Groq, Custom), making switching between them seamless. 🔑
• Improved Bluetooth support: Connecting to Bluetooth headsets is now faster and more reliable. 🎧
• Auto Enter: Added an option to automatically press Enter/Send after transcription. ⚡
• New Models: 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! 🎆

Dictate 3.1.0

This update adds an experimental feature called "auto formatting". If you have enabled auto formatting, you can use commands such as "new paragraph," "first," or "Henry with i" in the recording, which will then be automatically applied to your text. Processing will be a little bit slower, but the AI tries to process your instructions as good as possible. ✨

Furthermore, some bug fixes and theme changes have been made.

Dictate 3.0.0

This is the second major update for Dictate with plenty of improvements 🤩

• You no longer have to switch keyboards when you want to enter emojis or numbers. There are now dedicated buttons on the keyboard for that. 🥳
• All rewording instructions are now displayed. When you tap an instruction, the entire text is selected and revised automatically. If you selected only part of the text, only that part will be reworded. If you tap instructions during a recording, they are applied to your text right after the transcription. 🤖
• In the rewording instruction settings, you can now choose to have an instruction automatically applied after every transcription. That saves you plenty of unnecessary taps. 😌
• You can now easily export and import all of your instructions whenever you switch to a new device or want to share them. 📤
• The default style prompt is now sent to Whisper in the respective input language to avoid incorrect automatic translations. 💡
• Spanish and Portuguese have been added as app languages, along with an option to choose the language manually. 🇪🇸
• Some design changes and bug fixes have been made. ✨

Dictate 2.10.1

This is just a small update adding the ability to disable animations in the settings. Furthermore, Bluetooth support is disabled by default.

Dictate 2.10.0

This update brings some cool new features 🤩

• You can now record using the microphone of your headphones. So if you want to walk around more freely, this will help you a lot. 🎧
• If you tap, hold and swip to the left on the delete button, you can now select whole words for faster but still precise deletion. 🗑️
• You can now set a custom system prompt for rewording models in the settings. 💡
• Some design changes and bug fixes have been made. ✨

Dictate 2.9.1

This is just a small update adding the ability to automatically switch back to the previous keyboard after the transcription has finished. Simply do a long press on the recording button while recording. 🎉

Furthermore, I had to remove my PayPal reference from the settings to my GitHub page because of some Google restrictions. 🥲

Dictate 2.9.0

This update brings some cool new features 🤩

• You can now change the global theme and the accent color of the keyboard in the settings. So if you don't like light blue, this update is for you. 🎉
• OpenAI's new GPT-5 models have been added for rewording. 💡
• Some design changes and bug fixes have been made. ✨

Dictate 2.8.0

This update brings some cool new features 🤩

• In addition to OpenAI, you can now choose Groq as a provider for transcription and editing. Groq is generally free and significantly faster at processing audio recordings. To switch, simply scroll down to the settings — there you'll find a dedicated page where you can set the provider separately for each function, or even use your own service. 🎉
• The keyboard background now automatically switches to light or dark mode based on your system theme. 💡
• Proxy settings offer more flexibility now. ⚙️
• Dictate has a new app icon and improved Play Store listings. A big thank you to Jan for your work! 🤝
• Some design changes and bug fixes have been made. ✨

Dictate 2.7.2

This is just a small update adding the ability to configure a http proxy in the settings for all connections to OpenAI. 🛡️

Furthermore, some bugfixes have been made. ✨