LLM 0.37: new models, a cheaper default model, and better multimodal embeddings
Version 0.37 of the LLM CLI tool by Simon Willison adds the models gpt-6.1-sol and gpt-6.1-terra, a cheaper default model gpt-6-luna, and better support for multimodal embeddings (text, audio, image, video).
Simon Willison has released version 0.37 of his open-source CLI tool and Python library LLM, which serves as a unified interface for working with various language models. The release summarizes about two months of development and adds support for the new models gpt-6.1-sol and gpt-6.1-terra. The new command llm openai endpoint add is now available, which simplifies the configuration of additional models accessible via OpenAI-compatible endpoints.
The default model (for users who have not set a different one) has changed from gpt-5.6-luna to gpt-6-luna, which according to the author is half the price of its predecessor. Embedding support in the LLM tool has been expanded so that it can distinguish the type of binary content in embedded data — this is needed for models such as Gemini Embedding 2, which can create embeddings from text, audio, images, and video.
Willison also commented on his experience with contributions from AI-assisted developers: 12 authors contributed to the release, mostly with the help of AI tools, and according to the author, reviewing these contributions was not as burdensome as he expected. He describes the contributions as "bug reports with examples attached" — some he accepted directly, others he implemented in his own way.
Why it matters
For developers using the LLM tool, this means a direct saving (a cheaper default model) and an expansion of capabilities (new models, easier configuration of custom endpoints, embeddings for more types of content). Also interesting is the author's observation that an influx of AI-assisted contributions into an open-source project does not necessarily mean an increase in the burden of reviewing code.
Two audiences, two different impacts
What this means
For individuals
After updating, users of the LLM CLI tool can use the new models gpt-6.1-sol and gpt-6.1-terra, more easily add OpenAI-compatible endpoints with the command llm openai endpoint add, and pay less for the default model if they have not set their own default.
For a business
For companies using the open-source LLM tool as a layer for accessing various models, version 0.37 brings lower costs for the default model and broader support for multimodal embeddings, which can simplify internal tools without the need to write custom integrations for each provider.
DevelopmentCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.