GPT Engineer

Stale
GitHub Python MIT

Description

GPT Engineer is an AI tool that generates entire codebases based on natural language descriptions. Just describe what you want to build, the AI asks for clarification, and then builds it.

Key Features

  • Natural language to code — describe software requirements in natural language, AI generates entire codebases
  • Interactive improvement — pass improvement instructions to existing codebases, AI executes incremental changes
  • Custom pre-prompts — customize agent behavior by overriding the preprompts folder
  • Vision support — accepts UX/architecture diagrams as visual context for code generation
  • Multi-model compatibility — supports OpenAI, Anthropic, and open-source models like WizardCoder
  • Built-in benchmarking — bench tool for evaluating agents on APPS/MBPP datasets

Use Cases

💡 Rapid prototyping — shortest path from idea to runnable code
💡 Automated code improvement and refactoring for existing projects
💡 Evaluating and comparing different LLMs on code generation tasks
💡 Using visual input to assist generating frontend code from design diagrams
💡 Early experimentation platform for AI-assisted programming

Quick Start

pip install gpt-engineer
export OPENAI_API_KEY=your-key
mkdir my-project && echo 'Create a Flask REST API' > my-project/prompt
gpte my-project

Related Projects