> ## Documentation Index
> Fetch the complete documentation index at: https://plantis.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Release: llm 0.33

> The release of llm 0.33 includes upgrades to the OpenAI Python library, new key handling for embedding models, and enhanced template functionality.

<Info>
  **Original**: [Simon Willison](https://simonwillison.net/2026/Aug/22/llm/) · 22/08/2026
</Info>

## Summary

The release of llm 0.33 includes upgrades to the OpenAI Python library, new key handling for embedding models, and enhanced template functionality.

## Key Insights

> "The embedding models now use the same pattern for keys that regular LLM models do."
>
> — Discussing the new key handling in embedding models.

> "This unlocks a neat pattern where you can create templates that package a model with a set of default options."
>
> — Explaining the new template functionality.

> "Reasoning-capable Responses API models now support a reasoning\_summary option with auto, concise, and detailed values."
>
> — Highlighting the new features in the Responses API.

## Topics

* [openai](/kb/topics/openai)
* [python-library](/kb/topics/python-library)
* [embedding-models](/kb/topics/embedding-models)
* [llm](/kb/topics/llm)

***

## Full Article

My highlights from this release: Upgraded to the OpenAI Python library 3.x and switched the HTTP client dependency from httpx to httpx2. #1608, #1631 I shipped a quick 0.32.1 fix for this yesterday, but this is the more comprehensive fix. llm embed and llm embed-multi now accept --key. The Python EmbeddingModel.embed(), EmbeddingModel.embed\_multi(), Collection.embed() and Collection.embed\_multi() methods accept key= too, passing the resolved per-call key to embedding plugins without changing shared model state. Existing plugins that read self.key continue to work through a compatibility fallback. Thanks, ChrisJr404. #757, #1620 The embedding models now use the same pattern for keys that regular LLM models do. llm prompt -t/--template can 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. This unlocks a neat pattern where you can create templates that package a model with a set of default options: llm -m gpt-5.6-luna -o reasoning\_effort high --save lhigh llm "Generate an SVG of a pelican riding a bicycle" --save pelican # Combine and run the templates llm -t lhigh -t pelican Reasoning-capable Responses API models now support a reasoning\_summary option with auto, concise, and detailed values. This can be used with llm openai endpoint --responses. #1600 This is particularly useful for exercising different models that provide their own imitation of the OpenAI Responses API.

***

## Related Articles

<CardGroup cols={1}>
  <Card title="New release of LLM adds support for reasoning traces, OpenAI Responses, server-side tools, and smarter logging" icon="newspaper" href="/kb/articles/new-release-of-llm-adds-support-for-reasoning-traces-openai--996b94aa">
    Simon Willison · reference · 72% similar
  </Card>

  <Card title="Release: llm 0.32rc2" icon="newspaper" href="/kb/articles/release-llm-032rc2-fd4966fb">
    Simon Willison · reference · 71% similar
  </Card>

  <Card title="Release: llm 0.32a2" icon="newspaper" href="/kb/articles/release-llm-032a2-8db1edb7">
    Simon Willison · reference · 69% similar
  </Card>
</CardGroup>

***

<Note>
  Originally published at [https://simonwillison.net/2026/Aug/22/llm/](https://simonwillison.net/2026/Aug/22/llm/).
</Note>