ChatGPT CLI

Tiny ChatGPT terminal client.

• PyPI: https://pypi.org/project/cgpt-cli/
• License: MIT

Quickstart

make venv
source .venv/bin/activate        # macOS/Linux
# .\.venv\Scripts\Activate.ps1    # Windows

make compile
make install-dev

Authentication

Get an OpenAI API key from: 👉 https://platform.openai.com/account/api-keys

Store it via CLI:

cgpt login
cgpt where   # show config path

Config path:

• Linux/macOS: ~/.config/cgpt/config.toml
• Windows: %APPDATA%\cgpt\config.toml

If no config is found, OPENAI_API_KEY env var is used:

export OPENAI_API_KEY="your-key"   # macOS/Linux
setx OPENAI_API_KEY "your-key"     # Windows PowerShell

Configure Defaults

Use cgpt set to configure defaults like model and mode:

# Set a default model
cgpt set --model gpt-4o

# Set default mode (responses | chat)
cgpt set --mode responses

# Set both model and mode at once
cgpt set --model gpt-4o-mini --mode chat

Example config file:

api_key       = "sk-...redacted..."
default_model = "gpt-4o-mini"
default_mode  = "chat"

Development

pytest -q          # run tests
make sanity-check  # build + verify CLI

License

MIT © Mohammed Moosa Sharieff