LLM CLI
Command-line tool and Python library for running prompts across dozens of local and cloud-based language models from a single interface.
About
LLM is an open-source CLI tool built by Simon Willison and the Datasette project. You run it from the terminal to send prompts to models like GPT-4, Claude, Gemini, or locally installed Llama variants, and it logs every conversation and response to a local SQLite database for later review.
Beyond simple prompting, LLM supports generating and storing embeddings, extracting structured data from text and images, and granting models access to external tools. A plugin ecosystem extends support to additional providers. It is well suited for developers, data scientists, and power users who want to automate LLM tasks in scripts or explore multiple models without switching between apps.
Beyond simple prompting, LLM supports generating and storing embeddings, extracting structured data from text and images, and granting models access to external tools. A plugin ecosystem extends support to additional providers. It is well suited for developers, data scientists, and power users who want to automate LLM tasks in scripts or explore multiple models without switching between apps.