Version 5

On April 29, 2026, Simon Willison released two alpha versions of his popular llm CLI tool and Python library in rapid succession, marking a significant architectural overhaul. The headline change in 0.32a0 is the replacement of the previous prompt/response model with a message-sequence API that allows full prior conversations to be injected without requiring SQLite as an intermediary.[1] This is a backwards-compatible refactor, but a major one that reshapes how the library models both inputs and outputs.

The new streaming API is perhaps the most technically ambitious aspect of the release: responses are now exposed as typed event parts — text, tool_call_name, tool_call_args, and reasoning — enabling downstream consumers to handle the mixed-type outputs increasingly common from modern models like Claude.[1] Willison frames this as a direct response to the reality of contemporary LLMs: "Many of today's models return mixed types of content. A prompt run against Claude might return reasoning output, then text, then a JSON request for a tool call, then more text content." The CLI immediately leverages this by rendering reasoning/thinking tokens in a distinct color and routing them to stderr, keeping piped output clean.[1] A new to_dict/from_dict serialization mechanism rounds out the release, letting Python API users store and restore responses in any storage layer rather than being coupled to SQLite.[1] The 0.32a1 patch followed the same day, fixing a bug in which tool-calling conversations were not correctly reinflated from SQLite storage — a regression introduced by the architectural changes in 0.32a0.[2][3]

A third search cycle (May 1, 2026) has returned further confirmation that organic community reaction remains absent from the public record. The dominant signal is search noise: queries targeting "stable release" have surfaced dozens of entirely unrelated Twitter posts[4][5][6][7][8][9][10][11] matching the word "stable" with no connection to llm whatsoever. Mildly notable new signals include: an Instagram post specifically titled "LLM 0.32 just rewrote its core — and everything still ..."[12], extending social amplification to a new platform; a Bluesky post from Willison's own account[13] and a Fediverse post on "The LLM Python library support…"[14], both without extracted claims; and a cluster of simonw/llm GitHub issues on tool usage features[15][16][17] that provide ecosystem context for the kinds of problems the 0.32 refactor addresses, though none are explicitly about 0.32 migration. The llm-openai-plugin GitHub repository[18] and the official llm docs page for other models[19] are indexed without new content, and a general Willison plugins tag page[20] confirms ongoing plugin ecosystem activity without 0.32-specific commentary.

The broader pattern across three search cycles is now clear: the 0.32 alpha series has attracted aggregator and isolated social amplification (Twitter, Instagram, Bluesky, Fediverse) but has generated no substantive third-party analysis beyond the single explore.n1n.ai piece[21], no plugin author migration reports, and no Hacker News discussion thread. The llm GitHub issues on tool documentation[16] and tool-in-conversation continuity[15] indicate that tool-calling features were already a live concern in the project before the 0.32 release, lending further context to why the architectural refactor prioritized typed streaming event parts. Whether the stable release will surface community reaction that the alpha series has not remains the open question.