← All features

Everyday

Hands Free AI

Dictate long-form into a transcript you can edit, then send the whole thing to the chat at once.

Everyday · On the home screen by default · Asks for Microphone permission the first time

What it does

• Records continuously into a transcript that keeps growing, rather than doing one short dictation at a time. A large 'Start Recording' / 'Stop Recording' button and a running timer sit in the left sidebar.
• Lets you choose where speech recognition happens from an 'Input Source' menu — On-Device, OpenAI Whisper or Google Cloud. Cloud sources show a running cost estimate and a chunk count while you talk. If Apple's on-device recognition can't run because macOS Dictation is off, it falls back to Apple's server recognition and offers a one-time prompt with an 'Open Dictation Settings' button.
• Puts the transcript in an editable middle pane. It is locked while recording, with a note reading 'Recording in progress. Stop recording to edit transcript.', and becomes editable when you stop, so you can clean up what you said before anything is sent. 'Send to Chat' pushes it into the conversation on the right.
• Has an Auto-Send mode: turn on 'Auto-Send to Chat', type a word count or pick one of the presets (25, 50, 100, 250, 500, 1000), and it sends a chunk to the chat every time you've spoken that many more words, with a progress bar toward the next send and a count of chunks already sent.
• Saves sessions as named projects — Projects / New / Save buttons, an inline rename, and a 'Saved Sessions' sheet listing word count, message count and chunk count for each one.
• Recognises four spoken phrases that bypass the chat entirely: 'play music' and 'play podcast' navigate to those applets, 'stop listening' stops recording, and 'clear chat' (or 'start new conversation') wipes both the transcript and the conversation.

How to try it

1. On the home screen, click the Hands Free AI tile and allow microphone access the first time.
2. Click 'Start Recording' in the left sidebar and talk — the words build up in the Transcript pane in the middle.
3. Click 'Stop Recording', tidy the transcript if you want, then click 'Send to Chat'.
4. For a long dictation, switch on 'Auto-Send to Chat' first and pick a word count, so it sends a chunk every N words while you keep talking.

Works with

• Chat — the right-hand pane builds the same ChatViewModel and the same PersonaManager, so provider choice, attachments and personas behave as they do in Chat
• AI Music and Podcasts — the spoken phrases 'play music' and 'play podcast' navigate to those applets

Hands Free AI ships with mChatAI+. Native macOS app. Your own API keys, or fully on-device. Get mChatAI+

← Chat · Hub →