Optionally assign different language models to the OpenAI sync servers, format the chat area or activate individual sync servers for specific users only.
- Activate the option User-specific in the
Sync Server to make the AI accessible only to certain users/user groups. Then control the actual assignment of the Sync Server to specific users based on 
Sync Server accounts.
Example: A higher-performance and therefore more expensive AI model is allowed for individual users but not eveybody.
The instructions to the AI provide for different language styles per department (factual and brief for the finance department, elegant and detailed for the marketing department). - If you select Show response in real time, the AI response builds up step by step and the user quickly receives the first results. By default, the complete text is only output once the editing (including formatting) has been completed.
- In the Format for AI communication field, select a format from the picklist to better structure the chat history (bold, indented, etc.).
Depending on the provider, one format may work better than another. As AIs are getting better every day, it is not possible to make a definitive recommendation. The following formats are possible:- Text: plain text without formatting
- Text (format marker): Formatting/style possible
- MarkDown: Simple markup language for formatting text with normal text, is usually fast and well supported by AIs, but allows only little formatting.
- HTML: HTML formatting possible
- Restrict the AI models that are to be used in your company in the Released models field.
Example: gpt-5, gpt-5-mini - In the Standard OpenAI model field, select the default setting for the AI model to be used.
- Enter language preferences and general information for the AI in the Instruction to the AI field.
Example: "Use formal language. Answer the enquiry precisely and in detail. Mark the areas in the answer where you are not sure. - Enter important additional information about the sync server in the Description field.
