Open WebUI — Service Terms
Version 1.0 — April 2026
These Service Terms are an integral part of the NeoCloud Master Agreement and define the specific terms for the Managed Open WebUI service.
1. Service Description
The Managed Open WebUI service provides a dedicated AI chat interface deployed on a k0s cluster. The customer has full control over the deployment.
Included Features
- Dedicated Open WebUI instance per customer
- Managed PostgreSQL database with PgBouncer
- Integrated Qdrant vector database for RAG
- OpenAI-compatible inference endpoint connectivity
- Keycloak OIDC SSO integration
- Custom domain with TLS via cert-manager
- Velero cluster-level backups
- Full and differential PostgreSQL backups
2. Pricing Dimensions
| Dimension | Unit | Description |
|---|---|---|
| Cluster | Per cluster / month | Based on number of worker nodes and flavor |
| Storage | Per GB / month | Database and application storage |
| Inference | Per usage | Based on connected inference endpoint pricing |
Refer to the NeoCloud price list for current rates.
3. Customer Responsibilities
- Managing Open WebUI users and access controls
- Configuring inference endpoints
- Managing uploaded documents and RAG content
- Defining backup schedules in coordination with the Cloud Services team
4. Limitations
- GPU nodes for local inference (Ollama) are not supported in the current infrastructure
- Scaling is manual via the Cloud Services team
- Self-service provisioning is not supported in the current infrastructure
5. Definitions
| Term | Definition |
|---|---|
| RAG | Retrieval-Augmented Generation — a technique that augments LLM responses with information retrieved from uploaded documents |
| Inference Endpoint | An API endpoint that provides LLM capabilities (e.g., Nebul Inference, OpenAI) |
| SSO | Single Sign-On — authentication via an external identity provider |