WebAI To API

**This project is intended for research and educational purposes only.** Please use it responsibly and refrain from any commercial use.

WebAI-to-API

A FastAPI server that exposes Google Gemini (via browser cookies) as a local OpenAI-compatible API endpoint. No API key required — it reuses your existing browser session.

Compatible with any tool that supports the OpenAI API format: [Open WebUI](https://github.com/open-webui/open-webui), [Cursor](https://cursor.sh), [Continue](https://continue.dev), custom scripts, etc.


---

Deploy with Docker Compose

1. Create a folder and add two files

**`docker-compose.yml`**

services:
  web_ai:
    image: ghcr.io/leolionart/webai-to-api:latest
    container_name: web_ai_server
    restart: always
    ports:
      - "6969:6969"
    environment:
      - PYTHONPATH=/app/src
      - CONFIG_PATH=/app/data/config.conf
    volumes:
      - ./config.conf:/app/data/config.conf
    command: uvicorn app.main:app --host 0.0.0.0 --port 6969 --workers 1 --log-level info

**`config.conf`** (leave cookies empty for now)

[Browser]
name = chrome

[AI]
default_ai = gemini
default_model_gemini = gemini-3.0-flash

[Cookies]
gemini_cookie_1psid   =
gemini_cookie_1psidts =

[EnabledAI]
gemini = true

[Proxy]
http_proxy =

2. Start the server

docker compose up -d

3. Open the admin dashboard

Go to **`http://localhost:6969/admin`**

From there, paste your Gemini cookies and click **Connect** — no file editing needed.

---

Getting Gemini cookies

1. Open gemini.google.com and log in
2. Open DevTools (F12) → Network tab → refresh the page → click any request to gemini.google.com
3. Right-click the request → Copy → Copy as cURL
4. Paste the cURL command into the admin dashboard — it extracts the cookies automatically

Or manually: DevTools → **Application** → **Cookies** → copy `__Secure-1PSID` and `__Secur