Run Gemini 1.5 Flash from your CLI

cus why not

note: for any other llm, may have to replace request/response format (which here is uniqure for gemini)

Setup

1: clone repo to your directory

2: to avoid conflicts with OS python packages, consider running this in a Virtual Environment

python3 -m venv .venv
source .venv/bin/activate

3: Install Dependancy (Important)

pip install -r requirements.txt

4: create a .env file with your llm api key(check .env.example for format if unsure)

nano .env
API_KEY=your-llm-api-key-here

5: make the script executable

chmod +x gemini_cli.py

6: Rename and move to Global Path

mv gemini_cli.py llm
mkdir -p ~/.local/bin
mv llm ~/.local/bin/

7: Add ~/.local/bin to path

export PATH="$HOME/.local/bin:$PATH"
source ~/.bashrc  # or source ~/.zshrc

Use through CLI from anywhere now <3

llm "say yes if you agree kito is cool"