Use Your Own AI Model with InLoox
AI in InLoox Self-Hosted for the first time: The AI features used to be reserved for the cloud editions, because the default model runs in the InLoox Cloud only. With a model of your own, the full range of features is now available in a self-hosted installation as well:
- Generate project tasks from a short description
- Identify tasks in emails, both in the web app and in the Outlook add-in
- Have project descriptions written for you
To do so, register your own OpenAI-compatible endpoint in the new InLoox AI area of the Account Settings, which the cloud editions can do just as well if they prefer not to use the default model. Every AI feature then runs through that model, so you determine which provider is used and under which terms your data is processed. The same area is where you allow or block AI support for your account as a whole.
Connecting your own model is released as a beta feature in InLoox 26.5. The range of functions and the configuration options may still change.
The endpoint must provide an OpenAI-compatible API (Chat Completions). In the cloud editions it must be reachable from the internet, in InLoox Self-Hosted from the application server that runs InLoox. Configuring InLoox AI requires administrator permissions.
Help articles: InLoox AI ยท InLoox AI Assistant