jihchi / heygpt

A simple command-line interface for ChatGPT API written in Rust ⚡️

heygpt

A simple command-line interface for ChatGPT API.

• 🌟 Streaming output!
• 💡 One-shot mode to get a quick answer
• 🤖 Interactive mode to have a conversation

Demo (YouTube)

Quickstart

Install latest release version to $HOME/.cargo/bin/ via cargo:

cargo install heygpt

You'll need a OpenAI API key (you can get one here), and you'll need to export your API Key as an environment variable:

You can also set a OpenAI API base environment variable, just like openai-python

export OPENAI_API_KEY=<your api key>
# export OPENAI_API_BASE="https://api.openai.com/v1"

Then you can start an interactive conversation with ChatGPT:

heygpt

OR use the one-shot mode by providing a prompt:

heygpt how to record screen on mac

You can also compose prompt with bash tricks like

heygpt read the code and tell me what it is doing: $(cat src/main.rs)

heygpt read the code diff and write a commit message: $(git diff)

heygpt "please translate this poem to English:
> 床前明月光，
> 疑是地上霜。
> 举头望明月，
> 低头思故乡。"

You may even compose heygpt with other CLI tools via pipes:

$ echo "It's late. I should go to bed" | heygpt | cowsay
 ______________________________________
/ Goodnight! Sleep well and have sweet \
\ dreams.                              /
 --------------------------------------
        \   ^__^
         \  (oo)\_______
            (__)\       )\/\
                ||----w |
                ||     ||

Advanced

Commands in interactive mode

Enter \? to see available commands:

user => \?
Available commands:
  \?, \help: Show this help
  \b, \back: Retract and back to the last user message
  \h, \history: View current conversation history

Configuration file

heygpt will load configurations from $HOME/.heygpt.toml. You may also set API keys and base URL here. Example:

model = "gpt-4"
api_base_url = "https://some.openai.mirror/v1"
api_key = "your api key"