Know you're heard.
Ciao is the easiest way to talk to your computer. Your words appear while you are still speaking, so you see what it heard — and catch the one word that flips the meaning before it lands in your prompt.
Free and open source · runs on OpenAI or Google Gemini
- Live, not after the fact New words fade in bright and settle to grey, about 0.8 s after you say them.
- Hold a key, or just say "ciao" Push to talk on Right Ctrl, or turn on the wake word and start hands-free with your voice.
- Pastes where you started The text lands in the window you were typing in. Nothing is lost if the network drops.
Your key, your choice of engine
Ciao runs on your own API key, so you pay OpenAI or Google directly at their list price. No subscription, no middleman. Switch any time in Settings → Recognition (Gemini is desktop-only for now).
OpenAI
gpt-live-transcribe, one of the best speech-to-text models there is.
$1.02per hour of recording
Google Gemini
gemini-3.5-transcribe-live, which also applies spoken corrections: “at two, no, three” → “at three”.
~$0.54per hour of recording
- $0
- between dictations
- $0
- for Ciao itself