@ATheorell 
@similato87 
@TheoMcCabe 
@captivus What is this?
gpt-engineer is a CLI tool where you describe the software you want in a plain-text 'prompt' file, and an AI (OpenAI, Azure OpenAI, or Anthropic models) writes and executes the code for you; you can also ask it to implement improvements to existing code.
gpt-engineer is a CLI tool where you describe the software you want in a plain-text 'prompt' file, and an AI (OpenAI, Azure OpenAI, or Anthropic models) writes and executes the code for you; you can also ask it to implement improvements to existing code.Is this a library I import, or a standalone CLI tool?It is a standalone command-line tool installed with pip (or run via Docker) and invoked as 'gpte' against a project folder, not a library you import into your own code.
It is a standalone command-line tool installed with pip (or run via Docker) and invoked as 'gpte' against a project folder, not a library you import into your own code.How do I get it running?Install with 'python -m pip install gpt-engineer', create an empty project folder, add a file named 'prompt' containing your instructions, then run 'gpte projects/my-new-project' pointing at that folder.
Install with 'python -m pip install gpt-engineer', create an empty project folder, add a file named 'prompt' containing your instructions, then run 'gpte projects/my-new-project' pointing at that folder.Do I need an OpenAI API key to use it?Yes by default — export OPENAI_API_KEY or place it in a .env file; it also works with Azure OpenAI, Anthropic models, or local/open-source models like WizardCoder with a bit of extra setup.
Yes by default — export OPENAI_API_KEY or place it in a .env file; it also works with Azure OpenAI, Anthropic models, or local/open-source models like WizardCoder with a bit of extra setup.Is this project still actively maintained?No — the repository is archived, and the README itself points users toward gptengineer.app for the managed successor or 'aider' for a well-maintained, hackable CLI alternative.
No — the repository is archived, and the README itself points users toward gptengineer.app for the managed successor or 'aider' for a well-maintained, hackable CLI alternative.Does it work on Windows as well as macOS/Linux?Yes, there is a dedicated Windows README covering Windows-specific usage, in addition to the standard pip install and a Docker option.
Yes, there is a dedicated Windows README covering Windows-specific usage, in addition to the standard pip install and a Docker option.How do I improve existing code instead of starting a new project?Point gpt-engineer at a folder that already contains code, add a 'prompt' file describing the desired changes, and run 'gpte path/to/folder -i' to apply AI-driven improvements.
Point gpt-engineer at a folder that already contains code, add a 'prompt' file describing the desired changes, and run 'gpte path/to/folder -i' to apply AI-driven improvements.Can I benchmark my own coding agents with it?Yes, installing gpt-engineer also installs a 'bench' CLI for testing custom agent implementations against public datasets like APPS and MBPP.
Yes, installing gpt-engineer also installs a 'bench' CLI for testing custom agent implementations against public datasets like APPS and MBPP.Does it support image inputs for vision models?Yes, you can pass an image directory with the --image_directory flag alongside a vision-capable model like gpt-4-vision-preview, useful for giving the AI UX or architecture diagrams as context.
Yes, you can pass an image directory with the --image_directory flag alongside a vision-capable model like gpt-4-vision-preview, useful for giving the AI UX or architecture diagrams as context.What's the difference between this and gptengineer.app?gptengineer.app is the commercial, actively developed evolution with a UI for non-technical users and a git-controlled codebase, while this repository is the original open-source CLI experimentation tool, now archived.
gptengineer.app is the commercial, actively developed evolution with a UI for non-technical users and a git-controlled codebase, while this repository is the original open-source CLI experimentation tool, now archived.What are similar projects?gpt-pilot, plandex, self-operating-computer, Baize, aider, reverse-api-engineer, trae-agent, Agent-S
gpt-pilot, plandex, self-operating-computer, Baize, aider, reverse-api-engineer, trae-agent, Agent-SWhich topics is it in?codegen, autonomous-agent, coding-assistant, benchmarking, gpt-4, openai-api, vibe-coding, azure, code-generation, deprecated, anthropic, benchmark
codegen, autonomous-agent, coding-assistant, benchmarking, gpt-4, openai-api, vibe-coding, azure, code-generation, deprecated, anthropic, benchmark