llm 0.32a1

2026-04-29 · Source: Simon Willison's Weblog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, quick

Summary

The `llm` command-line tool, version 0.32a1, was released on April 29, 2026, providing access to large language models directly from the command line. This specific release addresses and fixes a critical bug identified in the preceding 0.32a0 version. The bug prevented tool-calling conversations from being correctly reinflated when stored and retrieved from SQLite databases. This update ensures the proper functioning of conversation history and tool integration within the `llm` utility, enhancing its reliability for users managing LLM interactions via the command line.

Key takeaway

For AI Engineers or developers using the `llm` command-line tool for large language model interactions, upgrading to version 0.32a1 is crucial. This update resolves a bug that could corrupt tool-calling conversations stored in SQLite, ensuring reliable conversation history and tool integration. Ensure your deployments reflect this patch to maintain data integrity and operational stability.

Key insights

The `llm` CLI tool's 0.32a1 release fixes a critical SQLite conversation reinflation bug.

In practice

• Update `llm` to 0.32a1 for bug fixes.
• Utilize `llm` for command-line LLM access.

Topics

• llm 0.32a1
• Command-line Interface
• Large Language Models
• Tool-calling
• SQLite

Code references

• simonw/llm

Best for: AI Engineer, Software Engineer, Prompt Engineer

Related on AIssential

• llm 0.32a0 — The `llm` CLI tool provides direct command-line access to large language models.
• LLM 0.32a0 is a major backwards-compatible refactor — LLM 0.32a0 refactors input to message sequences and output to typed streams, enhancing multi-modal and tool-use capabilities.
• Diagnosis Is Not Prescription: Linguistic Co-Adaptation Explains Patching Hazards in LLM Pipelines — Correcting LLM agent bottlenecks directly can degrade performance due to downstream modules' implicit adaptation to upstream errors.
• Alternatives for the EDIT tool of LLM agents — Token-efficient LLM agent editing can be achieved via line-specific checksum tags, improving reliability and speed.
• Why MCP Is Eating Your Context Window (and How Apideck CLI Fixes It) — Verbose JSON tool definitions in MCP rapidly consume LLM context windows, limiting conversational depth.
• The Sequence Opinion #852: The Bitter Lessons for Agentic Interfaces: A CLI for EVERYTHING — The article argues that the next evolution of agentic SaaS will shift from elaborate tool infrastructures and chat interfaces to providing Large Language Models (LLMs) with a complete Command Line Interface (CLI).

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Simon Willison's Weblog.