No description
Find a file
Zhizhin Vladislav f8ce67e721 Trim README to essentials
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-30 14:48:39 +07:00
src/owi_proxy Initial commit: OpenAI-compatible proxy for Open WebUI 2026-09-30 14:47:44 +07:00
.env.example Initial commit: OpenAI-compatible proxy for Open WebUI 2026-09-30 14:47:44 +07:00
.gitignore Initial commit: OpenAI-compatible proxy for Open WebUI 2026-09-30 14:47:44 +07:00
pyproject.toml Initial commit: OpenAI-compatible proxy for Open WebUI 2026-09-30 14:47:44 +07:00
README.md Trim README to essentials 2026-09-30 14:48:39 +07:00

owi-proxy

OpenAI-compatible proxy for Open WebUI: logs in via LDAP, refreshes the JWT on 401, and exposes /v1/* → <OWUI_URL>/api/*.

python3 -m venv .venv
source .venv/bin/activate.fish
pip install -e .
cp .env.example .env   # fill in credentials
owi-proxy

Then point any OpenAI client at http://127.0.0.1:8000/v1 with any api_key:

curl localhost:8000/v1/models

Configuration (.env)

Variable Description
OWUI_URL Open WebUI base URL
OWUI_USER, OWUI_PASSWORD LDAP credentials
OWUI_VERIFY_SSL true (certifi), false, or a path to a CA bundle, e.g. /etc/pki/tls/certs/ca-bundle.crt for a corporate CA
OWUI_CHAT_MODELS Comma-separated models (pipelines) that need a real chat_id; a temporary chat is created per request and deleted afterwards
PROXY_HOST, PROXY_PORT Listen address (default 127.0.0.1:8000)