Simon Willison — 2026-06-29#

Highlight#

Today’s standout piece is a hands-on exploration of Ornith-1.0, a newly released family of open-weights models specifically optimized for agentic coding. Simon tests its local execution capabilities and tool-calling proficiency, signaling another practical step forward for open-source AI developer tooling.

Posts#

Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding Simon goes hands-on with Ornith-1.0, a new MIT-licensed model family from DeepReinforce built on top of Gemma 4 and Qwen 3.5. Testing the 35B MoE variant locally via LM Studio, he finds it highly proficient at executing agent harnesses and running tool calls against a Datasette checkout. He highlights that the underlying models use clean Apache 2.0 licenses, successfully avoiding the “janky” terms of use that affected earlier Gemma models.

HTML table extractor Continuing his trend of building small, sharp utilities, Simon released a new paste-conversion tool that takes rich text from browsers and converts embedded tables into HTML, Markdown, CSV, TSV, or JSON. He also notes a recent update to his existing “Rich text to markdown” tool, which now features table support and an improved user interface.

Count the number of Safari tabs A classic, tiny “Today I Learned” (TIL) post sharing an AppleScript snippet for counting open browser tabs in Safari.

Project Pulse#

Simon continues to balance his focus between evaluating practical local LLM capabilities and building single-purpose web utilities. Exploring the boundaries of new agentic models while simultaneously releasing small paste-conversion tools perfectly encapsulates his hands-on approach to modern developer workflows.