{
  "id": 17838,
  "title": "llm 0.32rc2",
  "url": "https://urgent.news/2026/07/30/llm-0-32rc2",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-07-30T22:52:06.000Z",
  "source": {
    "name": "Simon Willison",
    "slug": "simon-willison",
    "url": "https://simonwillison.net/2026/Jul/30/llm-rc2/"
  },
  "original_language": "en",
  "account": "This is a breakdown of the latest update to LLM, version 0.32rc2. The most significant change is a dependency fix coupled with an update to the default model for users who haven't set their own preference. As of this update, GPT-5.6 Luna takes the place of GPT-4o mini as the default. Although Luna is a newer and more advanced model, it comes at a higher cost, priced at $0.20 for every million input tokens and $1.20 for every million output tokens. In contrast, GPT-4o mini remains a more budget-friendly option at $0.15 for input tokens and $0.60 for output tokens. Users can revert to the older model or switch to an even cheaper choice with the new GPT-5 nano, which costs $0.05 for input tokens and $0.40 for output tokens. Another noteworthy addition is the introduction of a new command for the llm openai endpoint. This feature allows users to run prompts, chats, and model listings against any OpenAI-compatible endpoint, without needing to first configure a model. Importantly, these calls are not logged, adding an extra layer of privacy. The llm openai endpoint command is described as a valuable addition to the LLM suite, addressing the lack of a straightforward CLI tool for testing prompts against various OpenAI Chat Completions imitation endpoints. Users with local LM Studio models can leverage this command without the need to install LLM, as demonstrated by a one-line example provided by Simon Willison, posted on July 30th, 2026. Finally, a sponsorship offer is included, inviting users to support the development by paying $10 per month for a curated digest of the month's most significant LLM developments.",
  "summary": "Release: llm 0.32rc2 Hot on the heels of RC1 , this fixes a dependency issue and also adds two neat new features: The default model for users who have not set their own default is now GPT-5.6 Luna . It was previously GPT-4o mini . Luna is a much better and more recent model, albeit slightly more expensive - $0.20 per million input tokens and $1.20 per million output tokens, compared to…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}