LLM is an Open Source command-line tool and Python library developed by Simon Willison for interacting with numerous Large Language Model providers such as OpenAI, Anthropic, and Google, as well as locally operated models. The current stable version 0.32 (released August 4, 2026) introduces visible Reasoning traces, support for the OpenAI Responses API, server-side tools (e.g., WebSearch, CodeInterpreter), and a new SQLite-based log storage. The project is open source and maintained on GitHub under the Apache-2.0 license.

Sources (2)