6 Python templates
Build Python apps with AI
Most AI app builders stop at JavaScript. Here a Python project is a real Python project: the framework's own layout, its own toolchain in the terminal, and a live preview of the running server.
The Python templates
Start opens the New Web App dialog with the template picked — you set the stack, design, model and MCP tools there.
- FastAPI
A typed async API with automatic OpenAPI docs at /docs.
- Flask
The small, explicit web framework — one file, no conventions to learn.
- Django
Batteries included — ORM, admin, auth and migrations, from django-admin.
- Streamlit
A data app from a plain script — widgets and charts, no frontend code.
- FastAPI AI SaaSAI SaaS
An AI SaaS dashboard: landing, Supabase sign-in, AI chat with history, MCP tools and knowledge search.
- Django AI SaaSAI SaaS
An AI SaaS dashboard: landing, Supabase sign-in, AI chat with history, MCP tools and knowledge search.
Every Python project gets
- A cloud sandbox with a live preview of the running app
- A real terminal with 15 coding CLIs — Claude Code, Codex, Cursor CLI, Gemini CLI and more
- AI on your own keys from 24 providers, billed by the provider at cost
- Its own GitHub repository from the first push
Where it can go
Django, FastAPI and Flask deploy to Vercel's Python runtime from the same window — use a managed Postgres rather than SQLite, since the deployed filesystem is read-only. Streamlit keeps a long-running server, so push it to GitHub and run it on a host that keeps a process alive.
Questions
Can I build a Python app with AI?
Yes. Pick one of 6 Python templates — Django, FastAPI, Flask and Streamlit — and describe what you want. The builder writes the code in a cloud sandbox with a live preview and a real terminal, using your own AI provider key.
Where can I deploy a Python app built here?
Django, FastAPI and Flask deploy to Vercel's Python runtime from the same window — use a managed Postgres rather than SQLite, since the deployed filesystem is read-only. Streamlit keeps a long-running server, so push it to GitHub and run it on a host that keeps a process alive.
What happens when I press Start?
You sign up — or, if you are already signed in, go straight on — and the New Web App dialog opens with that template picked. Set the name, stack, design, model and MCP tools there, then press Create.
Do I own the code?
Yes. The project is pushed to your own GitHub repository, and you can open a terminal on it at any point.
Start your Python app
7 days free, no card. Plans from $8 a month for the sandbox — AI is billed by your provider, at cost.
Pick a template