The short answer
Open DocAgent on the desktop, press Speak, say your request and press Stop listening. Your words appear in the box to edit before you send. In DocAgent voice settings, choose a voice and turn on Automatically read responses to hear the answers.
Works on: The DocFila desktop apps for macOS and Windows. Dictation is turned into text on your computer; online speech recognition is off unless you allow it. The assistant's reply is produced by AI on DocFila's servers and needs an account.
What do you need?
- A microphone and permission for DocFila to use it.
- The desktop app, signed in.
What are the steps?
- Open DocAgent. Press its key or use Ask DocAgent. The box says Ask DocAgent, or drop files here.
- Allow the microphone. The first time, your computer asks for microphone access. If it was denied, DocFila points to the privacy setting and offers Open privacy settings.
- Start speaking. Press Speak. The hint reads Listening, your words appear in the box to edit before sending.
- Finish. Press Stop listening. It reads Stopped listening, your words are in the box, ready to edit.
- Correct and send. Fix any word it misheard, then send the message like a typed one.
- Open the voice settings. In DocAgent voice, choose a Voice, use Preview voice and set the Speaking speed. Premium, Enhanced and Basic mark the installed voice quality.
- Turn on reading aloud. Switch on Voice playback, then Automatically read responses to hear each answer as it arrives. Press Esc or Stop speaking to silence it.
- Read a document aloud. While a document is open, DocAgent can read it aloud; the status shows Reading aloud.
What is worth knowing?
- Listening stops after two minutes. Your words stay in the box; press the microphone to go on.
- If DocFila says it did not hear anything, move closer to the microphone and try again.
- Natural voices come from your operating system. If only basic voices are installed, DocFila points you to the system settings to add better ones.
- Voice assistants, if you connect them, run searches and questions straight away but anything that could change files, send or sign still waits for you.
What DocFila does not do here
Dictation can fail for a language your computer cannot recognise offline. Then nothing is sent until you allow online speech recognition, which is off by default. Voice is a desktop feature; it is not part of the phone or web apps.
Where to go next
DocAgent Organize a folder with DocAgent DocAgent compared with chat assistants Where AI runs and what is sent
More ai how-to guides
How to draft a letter, request or notice with the AI Generator How to get a contract or letter explained in plain language How to organize and rename files with DocAgent, with approval and undo
Questions
Is what I say sent to a server?
By default, speech is turned into text on your computer only. If a language cannot be recognised there, DocFila asks before using an online speech service, and that option stays off until you switch it on.
Can DocAgent act on a spoken request?
Yes, like a typed one: the words go into the box, you send them, and any change to files appears as a plan you approve. Speaking does not skip the approval step.
How do I stop DocAgent talking?
Press Esc or Stop speaking while it reads. To silence it for good, turn off Voice playback in the DocAgent voice settings, which mutes DocAgent everywhere until you switch it back on.
Related DocFila tools
Keep going with the document workflow that matches this page. Open the tool you need, then finish the file in DocFila.