model roundup

Opus 4.7

2 items · started 2026-09-09 · closed 2026-09-13

  1. I think I may have figured out what is happening when Claude suddenly stops thinking in longer chats. I've seen other posts about Opus 4.7/4.8/5 using thinking for the first few turns and then suddenly giving instant, one-shot answers desp…

  2. AI coding agents often generate functionally correct but insecure code. To address this issue, we have open-sourced Security Cards, targeted security guidance for 80+ widely used libraries across 13 programming languages.

← all threads