Fetching from the wire…
Public story · 2026-08-10 · high
Simon Willison found out when his GitHub Actions job broke, a workflow built on free, keyless LLM access.
Why now: Willison posted about the retirement on August 9, more than a week after the July 30 cutoff, once his own workflow broke and forced the discovery.
GitHub retired its free Models API on July 30 with no announcement, and Simon Willison found out only when his Actions job broke, per his August 9 post.
The feature let code running in GitHub Actions call LLMs using the ambient GITHUB_TOKEN, no separate API key required. That was the base layer basically every "Continuous AI" demo ran on, and any CI job still calling GitHub Models that way is broken right now.
The error Willison hit called it a "scheduled retirement brownout," and GitHub hasn't said why it pulled the feature.
Willison's read is that usage from coding agents burned through the subsidized token budget faster than GitHub wanted to sustain, a pattern he says lines up exactly with the cost story KPMG has been telling. He migrated his own setup to a rate-capped OpenAI key.
GitHub pulled Models because keyless LLM access in CI couldn't survive agents calling models in a loop, and any platform still giving that away for free is running the same losing math. Watch whether OpenAI or Anthropic put rate caps on their own free-tier CI integrations next.
Willison posted about the retirement on August 9, more than a week after the July 30 cutoff, once his own workflow broke and forced the discovery.
Each link below shares sources, entities, or timing with this story.
Spotify's Portal team published Xirp on August 10: a vendor-neutral agentic development environment that manages concurrent sessions across Claude Code, Gemini CLI, and Codex, each session isolated in its own git worktree so dozens of agents can work the same codebase without...
Willison's August 2 roundup lays out "Open Weights and American AI Leadership" (July 24, Microsoft-shepherded, now 235 signatory companies including NVIDIA, Amazon, Y Combinator, the Linux Foundation, and OpenAI after initially abstaining); Anthropic's separate July 27 rebutta...
The number that reframes everything isn't ten. It's two thousand. OpenAI published "Ten advances in mathematics and theoretical computer science" on August 1, claiming an internal version of Astra produced new results on ten problems that had seen no progress on the main resul...
Promptwatch's tracking shows the share of ChatGPT search queries using site: sat at 0.3-0.5% for weeks, dipped to 0.15% on August 3-5, then jumped to 16-17% on August 8, two days after OpenAI said it was making GPT-5.6 Sol "more reliable with facts." Simon Willison Willison co...
An open-weight Chinese frontier model is now a dropdown option in Microsoft's coding product. That happened before anyone finished characterizing what the model does. GitHub's changelog dated August 6 makes Kimi K3 generally available across Copilot Pro, Pro+, Max, Business an...
Simon Willison published an analysis yesterday that I think will age well. His argument: Anthropic and OpenAI have found product-market fit, and it's not chatbots. It's coding agents. The numbers back him up. Anthropic's rumored Q2 2026 revenue hit $10.9B, up from $4B in Augus...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.