Chat AI Flexible

Chat AI Flexible provides access to the open-weight models hosted by GWDG in Göttingen.
It is designed for organizations that require monthly billing plans for flexible usage of AI models. Typical use cases include publicly accessible applications. The contract is valid for 12 months and can be extended upon request. The organization must also sign an Academic Cloud Basic contract.
Using the AI models via our web interface or API is optional.

Features of Chat AI Flexible

Priority access to cluster resources.
No restrictions on the list of available Open-Weight AI Models.
Full access to services including Chat AI service, Arcana/RAG, CoCo AI, Image AI, Voice AI, Protein AI when purchasing full access.
API keys for organizations, user groups, and individual users.
Organization-wide budget management and usage monitoring.
Usage tracked by token count according to our current pricing list.
Privileged access to the IDM for organizational administrators to manage users and budget limits.
Access to the SAIA Dashboard for organizational administrators to manage API keys, review usage, and monitor budgets.
Availability according to the Rate Limits.
Priority support via kisski-support@gwdg.de.

Contractual Terms

There are multiple variants available for the “Chat AI Flexible” contract:

Selection of Academic Cloud User Quota

The Academic Cloud Basic contract defines the maximum number of users from your organization who are allowed to register an Academic Cloud account. The Academic Cloud user quota is not included in the Chat AI Flexible contract price and must be ordered separately.
For the Chat AI Flexible contract, there is no quota corresponding to the Academic Cloud user quota. Instead, the organization defines usage limits, and monthly costs are calculated based on actual usage (see below).

Selection of Chat AI Flexible Limits

Organizations define monthly usage limits for AI models initially when signing the contract. These limits can only be changed via a support request from organizational administrators designated as authorized.

Soft Budget Limit obligatory: The organization defines a soft budget limit for monthly AI model usage. When the soft budget limit is reached, the organization is notified. The soft budget limit serves as a warning that monthly usage is approaching a certain value. The organization can then decide whether to continue using the AI models or reduce usage. They can also decide to increase the hard budget limit (if set) to prevent usage from being blocked in the current month.
Hard Budget Limit recommended: The organization can define a hard budget limit for monthly AI model usage. When the hard budget limit is reached, the organization is notified and AI model usage is blocked until the next month. The hard budget limit helps prevent unexpected costs for the organization.

Choice between Full Access and API-Only Access

Full Access: The organization can use the AI models via our web interface and/or via API.
API-Only Access: The organization can use the AI models only via API.
An Academic Cloud Basic contract is still required for this option to register the organization’s administrators.

Additional Documents for the Contract

Data Processing Agreement (AVV) with GWDG
Data Protection Officer and authorized persons
Technical responsible persons for the IDM
All contract documents will be provided with the offer.

Billing

Billing is based on actual AI model usage according to the current pricing list.
Billing starts at a minimum consumption of 1,000 € net.
If token consumption exceeds 1,000 € net, monthly billing begins.
The invoice amount equals the usage volume of the previous month (or 1,000 € if the usage volume threshold was reached), plus the usage volume of the previous month as a prepayment.
If actual usage in the current month is lower than the prepayment from the previous month, the difference is carried over to the next month.

Additional Conditions

The Terms of Use apply.