OmniAI CLI

Monday, July 15, 2024

OmniAI now includes a built-in CLI to simplify testing various LLMs directly from the console.

Installation

To use the OmniAI CLI, install the omniai gem along with any desired providers:

gem install omniai # required
gem install omniai-anthropic # optional
gem install omniai-google # optional
gem install omniai-mistral # optional
gem install omniai-openai # optional

Configuration

For each LLM, ensure your shell has the appropriate environment variable for the API key. For example, for OpenAI make sure ENV['OPENAI_API_KEY'] is present.

Usage

Once your shell is configured with the appropriate environment variables, an interactive chat can be launched by running:

omniai chat --provider="anthropic"

This opens an interactive session where any prompts are sent to the LLM and the response is streamed back:

Type "exit" or "quit" to leave.

# Who built you?
I was built by Anthropic.

# What is the capital of France?
The capital of France is Paris.

# What is the most popular language in France other than French?
The most popular language in France other than French is Arabic.

To run a single prompt (useful when chaining together shell commands), use:

omniai chat --provider="openai" --model="gpt-4" --temperature=0.9 "Tell me a joke"

Why don't scientists trust atoms? Because they make up everything!

For more detailed documentation simply run openai --help or browse the OmniAI CLI documentation.

This article originally appeared on https://workflow.ing/blog/articles/omniai-cli.