README.md

---
title: Visualisable AI Backend
emoji: 🧠
colorFrom: blue
colorTo: green
sdk: docker
pinned: false
short_description: LLM code generation with real-time trace visualization
---

# Visualisable.ai Backend Service

This is the backend service for Visualisable.ai, providing:

- Real-time model inference with trace extraction
- WebSocket streaming for live visualization
- REST API for model information and generation

## API Endpoints

- `GET /` - Health check
- `GET /health` - Detailed health status
- `GET /model/info` - Model architecture details
- `POST /generate` - Generate text with traces
- `WebSocket /ws` - Real-time trace streaming

## Configuration

Set the following secrets in your Space settings:

- `API_KEY` (optional) - API key for authentication

## Frontend

The frontend is deployed separately on Vercel. Connect it by setting the backend URL in your frontend environment variables.