llm 0.37

The latest 0.37 release of the LLM project—a command‑line tool and Python library for interacting with language models—brings two months of refinements and bug fixes. Key updates include the addition of the gpt‑6.1‑sol and gpt‑6.1‑terra models, and a new llm openai endpoint add command that streamlines configuration for OpenAI‑compatible models. The default model has also shifted to gpt‑6‑luna, which costs half as much as the previous gpt‑5.6‑luna.

Embedding support now recognizes various binary content types, enabling the use of models such as Gemini Embedding 2 that can process text, audio, images, and video. Simon Willison noted that, despite receiving many AI‑assisted pull requests, the review workload was lower than expected thanks to tools like Opus 5.5 and GPT‑6 Astra, which help prioritize the most valuable changes.

Twelve contributors—most of whom were assisted by AI—submitted fixes that were evaluated as bug reports with attached examples. Some were accepted unchanged, while others were resolved through custom implementations.

This update is significant because it demonstrates how AI tools can accelerate the development and maintenance of open‑source projects, reducing manual review effort and expanding users’ ability to work with diverse data types in text generation and analysis applications.

Read the original article on Simon Willison

This summary is an informational synthesis produced by dataqbs.com. All rights to the original content belong to its author and the cited media outlet. We act solely as curators of technology news and claim no authorship.