GPT-Engineer

GPT-Engineer is an archived open-source command-line agent that generates a whole codebase from a natural-language specification and then revises it on request. Anton Osika started the project in 2023, and its repository has since collected more than 55,000 GitHub stars.

Written in Python and installed from PyPI with python -m pip install gpt-engineer, it runs on Python 3.10 through 3.12. A project is a folder holding a file named prompt that describes the software in prose. Running gpte <project_dir> writes the code, and the -i flag switches to improve mode, which edits an existing codebase instead of starting fresh. Both modes read and write the same folder:

A project folder with a prompt feeds gpte in generate or improve mode, and the codebase lands back in the same folder.
Two Modes, One Project Folder

GPT-Engineer calls OpenAI models by default, and it can also target Azure OpenAI, Anthropic, or locally hosted open models. Custom preprompts override the agent’s built-in instructions, and a bundled bench command scores alternative agent implementations against the APPS and MBPP datasets.

The repository was archived in April 2026, and version 0.3.1 from June 2024 is the final release. Its commercial successor became Lovable.

Official website: github.com/gpt-engineer-org/gpt-engineer

A person in overalls pointing at a four-piece puzzle map labeled IDE, Cloud, CLI, and PR/Repo, with a map info legend beside it and a Python logo.

Tutorial

AI Coding Agents Guide: A Map of the Four Workflow Types

AI coding agents come in four types: IDE, terminal, PR, and cloud. Learn how each workflow fits into modern Python development.

intermediate ai

For additional information on related topics, take a look at the following resources:

By Martin Breuss • Updated Aug. 31, 2026
AI Coding Tools Share Feedback