A simple CLI chat tool designed for easy interaction with OpenAI's models. It also allows you to easily interact with Anthropic's Claude models and Groq's API.
- Basic CLI chat interface with OpenAI's LLM models.
- Web scraping capability for provided links (supports JavaScript scraping).
- YouTube scraping functionality for extracting video transcripts.
- The YouTube scraper either scrapes transcripts or downloads the video's audio and converts it to text using OpenAI's Whisper model though the Groq API.
- Multi-line support for easy copying and pasting into the CLI.
- Ability to generate images using OpenAI's image models.
- Uses CLImage to display images directly in the terminal.
- Supports both interactive and non-interactive chat modes.
- Basic support for other models on different platforms:
- Anthropic's Claude models via the
cla
command. - Groq's API through the
grq
command.
- Anthropic's Claude models via the
Click the image or visit the video HERE
Clone this repository, navigate to its directory, and run the following command to install or upgrade cha
:
pip3 install --upgrade .
-
Create a
.env
file in the root directory. -
Obtain your API keys:
- OpenAI API key: Get it here
- Anthropic API key: Get it here (for
cla
command) - Groq API key: Get it here (for
grq
command)
-
Add your keys to the
.env
file, using this format:# Replace YOUR_KEY_HERE with your respective API keys export OPENAI_API_KEY="YOUR_KEY_HERE" export ANTHROPIC_API_KEY="YOUR_KEY_HERE" export GROQ_API_KEY="YOUR_KEY_HERE"
-
To activate the environment variables, run:
source .env
To start the tool, execute one of the following:
# talk with OpenAI's models
cha
# talk with Anthropic's models
cla
# talk with Groq's supported models
grq
If you would like, you can use my configuration of cha
. To view my configuration, check out the cha.sh
file. To use it, add the content from the cha.sh
file to your shell's config file:
# zsh
cat cha.sh >> $HOME/.zshrc
# bash
cat cha.sh >> $HOME/.bashrc
Then, run my configuration of cha
:
chatgpt
1. For developing Cha, you can do the following. Install cha
in editable mode so that pip points to the source files of the cloned code:
pip install -e .
- To see and/or change hard-coded config variables/logic in Cha, checkout the
config.py
file.