Skip to content

Open WebUI: using the VIDAI Control Plane as the backend

Open WebUI's OpenAI connection settings accept a custom base URL and API key. Point them at the control plane and every user on your deployment chats with your VIDAI models.

TL;DR — Env vars

OPENAI_API_BASE_URL=https://your-vidai-server.example.com/v1
OPENAI_API_KEY=your-vidai-key

Set these on the Open WebUI container / process and restart. The models registered on the control plane appear in the model picker.

TL;DR — Admin Panel

  1. Admin Panel → Settings → Connections.
  2. Under OpenAI API, click + Add Connection.
  3. Set:
  4. API Base URL: https://your-vidai-server.example.com/v1
  5. API Key: your VIDAI API key
  6. Save. Refresh. Your VIDAI models appear in the picker.

Prerequisites

  • Open WebUI deployed (Docker, Kubernetes, or bare).
  • Control plane base URL and an API key from API Keys.
  • Models registered on the Models page.

Multiple connections

Add several OpenAI connections in the Admin Panel — Open WebUI merges the model lists in the picker. Useful for splitting keys per team:

  • VIDAI · Engineering → key A → cost-saver routing on
  • VIDAI · Compliance → key B → tight guardrails
  • VIDAI · Sales-ops → key C → deny-list on frontier models

Each key becomes a row on Chargeback.

Attribution

Same trade-off as any multi-user chat UI: the deployment authenticates with one (or several) shared keys. For per-user attribution, either give each team its own connection or use Open WebUI's built-in per-user API key support (Admin Panel → Users → per-user API key that the user pastes in Settings).

Verify it works

Log in, pick a VIDAI model, and send:

reply with exactly: ok

A row appears on Request Logs.

If something's off

Raise an issue at github.com/vidaiUK/vidai-quickstart/issues with the connection config (masking the key) and the Open WebUI version. We'll get it sorted.

Where to go next