{"slug": "llm-0-33", "title": "llm 0.33", "summary": "Simon Willison released llm 0.33, a command-line tool for running LLMs, which adds a --key option to llm embed and llm embed-multi commands and corresponding Python methods, allowing per-call keys for embedding plugins. The release also enables repeating -t/--template to combine templates, and adds a reasoning_summary option for reasoning-capable Responses API models. The update follows a quick 0.32.1 fix and includes contributions from ChrisJr404.", "body_md": "**Release:** [llm 0.33](https://github.com/simonw/llm/releases/tag/0.33)\n\nMy highlights from this release:\n\nI shipped a quick [0.32.1 fix](https://simonwillison.net/2026/Aug/21/llm/) for this yesterday, but this is the more comprehensive fix.\n\n`llm embed`\n\nand`llm embed-multi`\n\nnow accept`--key`\n\n. The Python`EmbeddingModel.embed()`\n\n,`EmbeddingModel.embed_multi()`\n\n,`Collection.embed()`\n\nand`Collection.embed_multi()`\n\nmethods accept`key=`\n\ntoo, passing the resolved per-call key to embedding plugins without changing shared model state. Existing plugins that read`self.key`\n\ncontinue to work through a compatibility fallback. Thanks,[ChrisJr404].[#757],[#1620]\n\nThe embedding models now use the same pattern for keys that regular LLM models do.\n\n`llm prompt -t/--template`\n\ncan now be repeated to combine templates in order. This allows model configuration and options from one template to be used with a prompt from another.\n\nThis unlocks a neat pattern where you can create templates that package a model with a set of default options:\n\n```\nllm -m gpt-5.6-luna -o reasoning_effort high --save lhigh\nllm \"Generate an SVG of a pelican riding a bicycle\" --save pelican\n# Combine and run the templates\nllm -t lhigh -t pelican\n```\n\n- Reasoning-capable Responses API models now support a\n`reasoning_summary`\n\noption with`auto`\n\n,`concise`\n\n, and`detailed`\n\nvalues. This can be used with[llm openai endpoint --responses].[#1600]\n\nThis is particularly useful for exercising different models that provide their own imitation of the OpenAI Responses API.\n\nTags: [annotated-release-notes](https://simonwillison.net/tags/annotated-release-notes), [llm](https://simonwillison.net/tags/llm)", "url": "https://wpnews.pro/news/llm-0-33", "canonical_source": "https://simonwillison.net/2026/Aug/22/llm/", "published_at": "2026-08-22 17:01:16+00:00", "updated_at": "2026-08-22 18:13:06.920010+00:00", "lang": "en", "topics": ["developer-tools", "artificial-intelligence", "large-language-models"], "entities": ["Simon Willison", "llm", "ChrisJr404", "OpenAI"], "alternates": {"html": "https://wpnews.pro/news/llm-0-33", "markdown": "https://wpnews.pro/news/llm-0-33.md", "text": "https://wpnews.pro/news/llm-0-33.txt", "jsonld": "https://wpnews.pro/news/llm-0-33.jsonld"}}