Adjust reasoning and speed: fast mode and auxiliary models
Use AI Model to adjust reasoning effort, fast mode and auxiliary tasks. Faster replies may cost more or consume your subscription allowance faster.
Where to go
Open Mate Settings → choose the mate → AI Model.
Supported models in Model order show Effort. Fast mode appears only when the main model, direct-provider route and mate version support it. It may also be absent while data is loading. Check model status or contact support if needed.
Reasoning effort
- Save any model-order edits first.
- Choose effort beside the model. Levels may include Off, Low, Medium, High, Extra high and Max.
*marks the default. - Wait for processing and check the selected value and success notice.
Changes apply from the next conversation. Higher effort usually takes more time and costs more. Unsupported levels are lowered; check the returned value. A note identifies settings shared by models with the same name.
Four fast modes
N is the number of seconds displayed in your dashboard.
| Mode | Scope |
|---|---|
| Off | No speed request. |
| First N seconds of the first turn | Speeds up the opening of a new conversation. |
| First N seconds of every turn | Speeds up each reply’s opening. |
| Always fast | Throughout supported messaging-app conversations; dashboard chats still use the first-N-seconds window each turn. |
Scheduled work is not sped up. A conversation-specific fast setting in a messaging app may take precedence over the dashboard’s global setting.
Choose a mode, read and confirm its cost, then wait for the updated value. It applies from the next message. Select Off to stop and check the confirmation.
Faster processing uses your own provider account’s billing or subscription allowance. Actual prices and limits are set by that provider.
Changing the main model attempts to turn fast mode off. If the dashboard could not confirm that, check the setting and choose Off yourself.
Auxiliary tasks
Model for auxiliary tasks covers images, conversation compression and titles. Usually Same as the main model (automatic) is enough.
- Select Change.
- Choose automatic or a backup already in Model order.
- Select Apply and confirm.
- Check the current setting and success notice.
Add another model to Model order first. Removing a selected backup returns auxiliary tasks to the main model. Contact support for custom-endpoint settings that cannot be changed here.
If a change fails
Use the returned value as the actual setting. Save the model order first if it is still a draft; reselect if the model was removed. Contact support for repeated failures. Check AI usage to review estimates.
fast mode faster replies reasoning effort thinking level speed auxiliary model
Need a hand? Contact support · Cannot sign in? Email us