Skip to content
Open Source

Your First Production AI App
in 30 Minutes

Choose from 5 AI frameworks, configure 75+ options, and deploy. FastAPI + Next.js with auth, streaming, and infrastructure — ready out of the box.

Configure & Download View on GitHub 1.5k
Terminal

Works with your stack

Pydantic AI
LangChain
LangGraph
CrewAI
DeepAgents
FastAPI
Next.js
Python
TypeScript
PostgreSQL
MongoDB
Redis
Docker
Kubernetes
Sentry
Prometheus
Logfire
Tailwind CSS
Pydantic AI
LangChain
LangGraph
CrewAI
DeepAgents
FastAPI
Next.js
Python
TypeScript
PostgreSQL
MongoDB
Redis
Docker
Kubernetes
Sentry
Prometheus
Logfire
Tailwind CSS

Why use this?

Without the template

  • ✗ Days setting up FastAPI + Next.js boilerplate
  • ✗ Manual AI framework integration
  • ✗ No WebSocket streaming out of the box
  • ✗ Docker and deployment configured from scratch
  • ✗ No observability until you add it yourself

With the template

  • ✓ Production-ready project in minutes
  • ✓ 5 AI frameworks pre-integrated and tested
  • ✓ WebSocket streaming with tool call visualization
  • ✓ Docker, Kubernetes, and CI/CD included
  • ✓ Logfire, Sentry, and Prometheus built-in

How It Works

From zero to production in four steps

1

Pick Your Stack

Choose from 5 AI frameworks, 3 frontends, and 75+ configuration options via the web configurator or CLI.

2

Generate & Scaffold

One command generates a fully configured project — Docker, CI/CD, auth, database, and your selected AI stack.

3

Build Your Agent

Focus on your AI logic. The template handles routing, streaming, state management, and deployment boilerplate.

4

Ship to Production

Deploy with docker compose up. Monitoring, logging, and health checks are pre-configured out of the box.

Everything you need

From database to deployment, configured in one command.

5 AI Frameworks

Pydantic AI, LangChain, LangGraph, CrewAI, and DeepAgents with WebSocket streaming.

75+ Options

Database, auth, background tasks, caching, rate limiting, file storage, webhooks, and more.

3 Presets

Minimal, Production, or AI Agent. One command, zero configuration.

Full-Stack

FastAPI backend with Next.js 15 frontend. TypeScript, App Router, Tailwind CSS.

Production Ready

Docker, reverse proxy, Kubernetes, CI/CD, .env generation out of the box.

Observable

Logfire, Sentry, and Prometheus integration. Know what your agents are doing.

How does it compare?

See how it stacks up against alternatives.

Feature AI Agent Template Manual Setup Other Boilerplates
AI Agent Integration ✓ ✗ ✗
Multi-Database Support ✓ ✗ Partial
Auth Built-in ✓ ✗ Partial
WebSocket Streaming ✓ ✗ ✗
Web Configurator ✓ ✗ ✗
Type-Safe Code ✓ Varies Varies
Observability ✓ ✗ ✗

Start with a preset

Or customize every option in the configurator.

Minimal

API-only, no database, no auth. Perfect for prototyping.

  • — No database
  • — No authentication
  • — No Docker
  • — Basic FastAPI
Use Minimal

Production

Recommended

Full production setup. Deploy-ready from day one.

  • — PostgreSQL + JWT auth
  • — Redis + Caching
  • — Sentry + Prometheus
  • — Docker + Kubernetes
  • — GitHub Actions CI
Use Production

AI Agent

Most Popular

AI-powered app with streaming and persistence.

  • — Pydantic AI (default)
  • — WebSocket streaming
  • — Conversation history
  • — PostgreSQL + Redis
  • — Next.js frontend
Use AI Agent

Get started

1

Install

$ pip install fastapi-fullstack
Click to copy
2

Create

$ fastapi-fullstack new
Click to copy

Or with a preset: fastapi-fullstack create my_app --preset ai-agent

3

Run

$ cd my_app && docker compose up
Click to copy
Or use the web configurator

Frequently Asked Questions

Answers to common questions about the Full-Stack AI Agent Template, Pydantic DeepAgents, and Logfire integration. Find guides and troubleshooting tips.

What is the Full-Stack AI Agent Template?
An open-source project generator that creates production-ready AI/LLM applications with a FastAPI backend and Next.js frontend. One CLI command or the web configurator generates a complete project with your choice of 5 AI frameworks, 3 databases, 4 vector stores, and 75+ configuration options. The project has 1,730+ GitHub stars and 830K+ downloads.
Which AI framework should I choose?
The template supports 5 AI frameworks. Choose Pydantic AI for type-safe, production-grade agents with Logfire observability. Pick LangChain for its 700+ integrations. Use LangGraph for complex multi-step workflows with state management. Try CrewAI for multi-agent collaboration. Select DeepAgents for autonomous agents with planning and human-in-the-loop approval.
Can I switch AI frameworks after generating a project?
Yes. Regenerate the project with a different --ai-framework flag. Your custom code outside the generated agent module is preserved if you use version control. The web configurator also lets you export your configuration as JSON and re-import it later. The agent module is isolated to ~150 lines, so switching frameworks requires no changes to your routes, auth, or frontend code.
Is the template free to use?
Yes, completely free. The template is MIT licensed — use it for personal and commercial projects without restrictions. No premium tiers, no usage limits, no sign-up required. With 1,730+ GitHub stars and an active community, it is one of the most popular open-source AI app generators available.
Which database should I use?
PostgreSQL is recommended for production — it supports the admin panel, conversation persistence, and full SQLAlchemy/SQLModel ORM features. Use MongoDB for document-oriented workloads. SQLite is great for development and small deployments with zero configuration. Choose 'None' for stateless API-only services. The template also supports 4 vector stores (Qdrant, ChromaDB, Pinecone, pgvector) for RAG workflows.
See all questions

Frequently Asked Questions

What is the Full-Stack AI Agent Template?

Open-source project generator for production AI apps with FastAPI + Next.js. 5 AI frameworks, 75+ options, 3 presets. 1,730+ stars.

Which AI frameworks does it support?

Pydantic AI, LangChain, LangGraph, CrewAI, and DeepAgents. Switch frameworks without changing your backend or frontend code.

Does it include RAG?

Yes, v0.2.2 includes a full RAG pipeline: 4 vector stores (Milvus, Qdrant, ChromaDB, pgvector), hybrid search with BM25 + RRF, Cohere/CrossEncoder reranking, 4 embedding providers.

Is it free?

Yes, MIT licensed. No signup required. Use the CLI (pip install fastapi-fullstack) or the web configurator.

Can I use it in production?

Yes, battle-tested across 30+ production deployments. Includes Docker, Kubernetes, CI/CD, auth, observability, and background tasks out of the box.

Ready to ship your AI app?

Pick your frameworks, generate a production-ready project, and deploy. 75+ options, one command, zero config debt.

Configure & Download