Auto mode
Let EchoChat pick a fitting model for each request instead of choosing one yourself, while staying within your organization's model and privacy rules.
What Auto mode does
Auto mode (Beta) lets EchoChat choose a fitting model for each message, so you do not have to pick one yourself. Simple requests go to a fast, efficient model; more demanding ones are handled by a stronger model. The goal is a hands-off experience that stays quick and cost-aware without you thinking about model names.
How it decides
- It looks at your message and estimates how demanding the task is.
- Short, everyday requests use a fast model. Requests that need real reasoning, careful analysis, or code get a stronger one.
- Every message is assessed, and the level can only go up within a conversation - so answers stay consistent, and a chat that starts casually still gets a stronger model the moment a harder question arrives. Regenerating a reply asks for a stronger model.
Which models Auto can choose
- Only models your organization allows. Auto never selects something an admin has switched off.
- It stays within your Zero Data Retention policy: when ZDR is active, Auto only uses models that meet it.
- Admins can steer it further, for example to prefer EU-based providers or to cap it at a maximum level.
Turning it on
Open the model selector and choose Auto at the top. It applies to text chats; image and video generation keep their own selectors. You can switch back to a specific model at any time.
Good to know
Auto mode is in beta. It is designed to be reliable for everyday work, and it always respects your organization's model and privacy settings. If a particular answer would be better on a specific model, you can always select that model yourself for that chat.