Voice Input

Every prompt input in the app has a microphone button. Click it to dictate instead of typing.

Where to use it

• Generate Image panel
• Generate Video panel
• Agent dock chat input

The microphone icon sits next to the prompt textbox or send button.

How it works

1. Click the microphone. The app asks for microphone permission on first use (system prompt).
2. Speak. A live waveform indicates recording.
3. Click the microphone again, or pause for a moment, to stop recording.
4. The audio is sent for transcription. The transcribed text replaces or extends the prompt box content.

Transcription provider

Transcription uses the configured AI provider — usually OpenAI Whisper (if an OpenAI key is set) or Google Gemini (if a Google key is set). The first available one is used.

If neither key is set, the microphone button is hidden.

Cost

Whisper and Gemini transcription cost a fraction of a cent per minute of audio. The cost is included in your provider’s usage like any other API call.

Privacy

• The audio file is sent only to the configured provider.
• No audio is stored on disk after transcription.
• The microphone permission can be revoked in macOS System Settings → Privacy & Security → Microphone or Windows Settings → Privacy → Microphone.

Tips

• Speak naturally — modern transcription handles full sentences better than keyword-by-keyword dictation.
• Switch language — Whisper and Gemini both auto-detect; you can dictate in any major language.
• Punctuation — say “comma”, “period”, “new paragraph” and the transcriber inserts the punctuation literally.

Where to go next

• Image Generation
• Audio Transcription — separate endpoint for transcribing audio elements.