| title: glmgate – Free GLM Reverse Proxy | |
| emoji: 🦎 | |
| colorFrom: gray | |
| colorTo: green | |
| sdk: docker | |
| app_port: 7860 | |
| # glmgate | |
| A **fully‑automated** OpenAI‑compatible reverse proxy for the free GLM chat at <https://chat.z.ai>. | |
| - No API keys required – accounts are created on‑the‑fly using temporary email addresses. | |
| - Captcha handling is done with Playwright (headful, Xvfb). | |
| - Tokens are refreshed automatically; dead accounts are replaced without user involvement. | |
| - Comes with a **Svelte SPA dashboard** to monitor accounts, models, request rate and live logs. | |
| - Deploys on a **free Hugging Face Space** (CPU‑only, port 7860). | |
| ## Quick start (local dev) | |
| ```bash | |
| # Build frontend | |
| cd frontend && npm ci && npm run build | |
| # Install backend deps and run | |
| cd .. | |
| pip install -r requirements.txt | |
| uvicorn main:app --host 0.0.0.0 --port 7860 | |
| ``` | |
| The dashboard will be reachable at <http://localhost:7860>. | |
| ## Deploy to Hugging Face Spaces | |
| The repository already contains a multi‑stage Dockerfile that builds the frontend and bundles the FastAPI backend with Playwright. | |
| Just create a new Space, point it at this repo and enable the **Docker** runtime. The Space will automatically build and start. | |
| --- | |
| *Built with love by DEV & Fickito.* | |