During the beta testing phase, you can use Luna for free and get a feel for how it works. In its current version, Luna has already been extensively tested by our staff and by
ELITECAD Insiders. This means you’ll receive an AI integration that has already been pre-tested and fine-tuned. Nevertheless, we’re constantly expanding the features, learning from
user feedback, and working out the details for fair packages that you’ll be able to add as an optional AI module to ELITECAD in the future. The beta testing phase is designed to be long-term yet will be limited in its duration. The beta testing phase is expected to end mid-2027 (subject to change) with advance notice provided in good time so you can switch to the future AI module if you want to. Interested in more than just the beta before then? Fill out
this form and perhaps we can make something happen...
During the beta testing phase, we will cover the costs of the underlying AI model for all users. However, we have set budget and token limits that apply to test and training licences, as well as to our customer licences. In each individual project, a maximum of 1,000,000 tokens per licence and calendar month are allocated, of which up to 25% may be used for image generation. If the limits are exceeded or the system is misused, we may temporarily suspend the AI integration for all licences or permanently suspend it for specific licences. We also reserve the right to change the budget and token limits at any time after the start of the beta testing phase. The traffic light colours used to indicate your token consumption per prompt in the AI chat window show you how many of your assigned monthly free tokens are left (red = <25%, yellow = 25–50%, green = 50–100%). The token balance is reset at the start of each calendar month.
The AI model used is pre-selected during the beta testing phase to streamline the process and therefore cannot (yet) be chosen individually. Technically, Luna is based in the test phase on an API connection to OpenAI, and data processing takes place within the framework of a data processing agreement. Transmissions of queries to the AI model are pseudonymised, meaning they are tagged with identification numbers or other identifiers rather than personal data or office data as the sender. During the test phase, transmission and processing take place in joint projects according to licence type. Personalisation of Luna is therefore not yet possible at this stage; the context is lost after each programme restart. This means that every time ELITECAD is restarted during the beta test phase, the knowledge from previous conversations is reset. You can find a small, temporary workaround for this issue under ‘Can I personalise the AI assistant in ELITECAD?’.
To ensure smooth operation and the further development of the service, a usage log is also created on secure XEOMETRIC servers within the EU whenever Luna is used. This log includes licensing data, which makes it possible for XEOMETRIC to identify users. This is currently also important for preventing misuse of the free trial period.
As described in the security information prior to using the AI assistant Luna, which users must confirm, it is the responsibility of the AI user to ensure at all times that they do not enter any sensitive, personal or confidential content directly into the chat window. Such entries are neither pseudonymised nor anonymised by XEOMETRIC.