> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://help.playability.gg/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# Set Up Voice Commands

Voice commands can trigger compatible gamepad, keyboard, mouse, and PlayAbility actions. PlayAbility recognizes single words and multi-word phrases without regard to letter case.

## Select the microphone and language

1. Open **Settings** > **General**.
2. Under **Microphone**, select the device you want to use.
3. Under **Voice Recognition**, select the **Voice Recognition Language** that matches what you will speak.
4. If the language model is not installed, wait for the download to finish.
5. Confirm the selected language says that its model is ready for use.

Changing the language can require another model download. Keep PlayAbility open and connected to the internet until that download completes.

## Create a voice-command trigger

1. Open the current profile in **Mapping**.
2. Add the action you want to control, or find an existing action without a trigger.
3. Select **Select a trigger**.
4. Select **Create a new trigger +**.
5. Under **How do you want to control it?**, select **Voice**.
6. Under **Enter Word or Phrase**, type the command.
7. Select **Add** or press Enter.
8. Choose an **Action Mode**:
   - **Short Press** is a good default for a button, key, click, or one-time action.
   - **Long Press**, **Double Press**, and **Triple Press** send the corresponding fixed press pattern.
   - **Toggle** switches the output on or off each time the command is recognized.
9. Select **Save**.

Use short, distinct commands that the selected language model can recognize consistently. Examples include “jump” for a keyboard Space action, “reload” for a gamepad button, or “pause game” for a Start action.

## Verify Voice Detection

1. Start **Voice Detection** beside the webcam preview.
2. Wait until PlayAbility reports that it is listening for commands.
3. Make sure **Main Output** is enabled.
4. Say the command clearly at a normal volume.
5. Confirm the recognized word or phrase appears in the voice feedback.
6. Confirm the mapping and its output activate.
7. Repeat the command several times and test it in the target game or app.

## Troubleshooting and limitations

### Voice Detection remains on Loading or Downloading

Wait for the selected speech model to finish downloading. If the model remains unavailable, return to **Settings** > **General**, confirm the language, and try **Clear voice model cache** before starting Voice Detection again.

### Nothing is recognized

Confirm the correct microphone is selected and Windows has granted microphone access. Make sure **Voice Detection** is running, then speak briefly to warm up the model.

### Other words are recognized, but the command is not

Use a clearer or more distinct command, verify the selected recognition language, and avoid words that sound very similar. For a phrase, PlayAbility must recognize its words in the same order.

### The command is recognized but no action occurs

Confirm the intended profile is current, the voice trigger is saved on the action, and **Main Output** is enabled. Check the mapping's **Action Mode** and the target application's focus.

### A command fires unexpectedly

Choose a less common word or a multi-word phrase. Reduce background speech and avoid command words that are likely to occur in conversation or game audio reaching the microphone.

### A spoken command cannot be held like a physical button

Voice is an event rather than a continuously held input, so it does not use **Default** hold behavior. Use a press mode for a fixed press or **Toggle** for an output that must remain active until the command is spoken again.

### Repeating a word does not retrigger as expected

Pause briefly between separate commands so the recognizer treats them as separate speech events. Watch the voice feedback to confirm how the model transcribed the speech.

## Related articles

- [Build and Edit a Mapping](https://help.playability.gg/en/article/build-and-edit-a-mapping-u0q0bs/)
- [Map Inputs to Keyboard Actions](https://help.playability.gg/en/article/map-inputs-to-keyboard-actions-b9pj1d/)
- [Action Modes and Advanced Trigger Logic](https://help.playability.gg/en/article/action-modes-and-advanced-trigger-logic-oiuj2g/)