← all metas

Thursday, April 16, 2026

Opus 4.7 drops with self-verification and xhigh effort — builders integrate immediately, memes about un-nerfed models follow

Claude Opus 4.7 launch with agentic features

Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.

Also dominant that day

  • @OpenAI OpenAI Codex Mac computer use & agent upgrades — Codex gains Mac computer use, memory, and background agents — agentic workflow hype meets Windows-user salt
  • @RohOnChain Stanford LLM fundamentals lecture goes viral — A Stanford lecture on LLM internals goes viral — practitioners call it better than most lab onboarding