Open WebUI

Open WebUI adds ozeye as an OpenAI API connection and fetches the model list from /v1/models, so every ozeye chat model shows up in the model picker.

In the admin settings

  1. Open Admin Panel → Settings → Connections.
  2. Under OpenAI API, add a connection with URL https://ozeye.ai/v1 and your oz_… key.
  3. Click Verify connection (saving alone does not test it), then save.
  4. Optional: set a Prefix ID such as ozeye to show models as ozeye.glm-5.3 next to other providers. The prefix is removed before the request reaches ozeye.

With Docker

docker run -d -p 3000:8080 \
  -e OPENAI_API_BASE_URL=https://ozeye.ai/v1 \
  -e OPENAI_API_KEY=oz_your_key_here \
  -v open-webui:/app/backend/data \
  --name open-webui --restart always \
  ghcr.io/open-webui/open-webui:main
# ozeye next to another provider: same order in both lists
OPENAI_API_BASE_URLS="https://api.openai.com/v1;https://ozeye.ai/v1"
OPENAI_API_KEYS="sk-...;oz_your_key_here"
Environment variables only apply on the first start. After that, Open WebUI keeps the connection settings in its database, so change them in the admin settings (or set ENABLE_PERSISTENT_CONFIG=False).

Good to know

  • Filling in Model IDs replaces the fetched list completely; leave it empty to get every ozeye model.
  • Requests go from your Open WebUI server to ozeye. ozeye does not store the prompts or responses; your chat history stays in your own Open WebUI database.
  • Web search, image generation and speech in Open WebUI use their own settings and providers; this connection covers chat.