# Simon Willison — The Last Six Months in LLMs, in Five Minutes

> A five-minute tour of what changed in large language models between late 2025 and May 2026.

- Org: Simon Willison
- Date: 2026-05-19
- Category: article
- Source: https://simonwillison.net/2026/May/19/5-minute-llms/
- Canonical page: https://ai-tldr.dev/releases/simonw-six-months-llms/

## In short

Simon Willison's PyCon US 2026 lightning talk recaps six months of LLM progress: frontier models trading the top spot, coding agents going from often-work to mostly-work, and open-weight models now running well on laptops.

## What it is

A written version of Simon Willison's lightning talk at PyCon US 2026. Willison is the creator of Datasette and the LLM CLI and writes one of the most-followed running commentaries on LLM developments.

## How it works

The post walks through three themes: multiple frontier labs (Anthropic, OpenAI, Google) alternating as the best model, coding agents maturing from unreliable to daily-driver tools via reinforcement learning, and open-weight models like Gemma 4, GLM-5.1, and Qwen becoming strong enough to run on consumer laptops. It uses his recurring 'pelican riding a bicycle' SVG benchmark to compare models.

## Why it matters

It is a fast, opinionated catch-up for anyone who tuned out of AI news for six months, from a source known for verifying claims and avoiding hype.

## Who it's for

developers catching up on AI

## Links

- [Blog Post](https://simonwillison.net/2026/May/19/5-minute-llms/)
- [HN Discussion](https://news.ycombinator.com/item?id=48188183)
