A while back I built a small tool called SDKProof. it checks how well an AI coding agent writes an...
Claude Opus 5 achieved 100% on mid-2025 SDKs (Vercel AI SDK 7, Zod 4) but only ~80% on late-2025 ones (Prisma 7, Next.js 16) because training data lags major API changes. For teams building with cutting-edge frameworks, AI-generated code still requires runtime validation; the model-API knowledge gap resets with each major release.
A while back I built a small tool called SDKProof. it checks how well an AI coding agent writes an SDK's current API — the stuff that changed in the last major, that the model tends to get wrong because it learned the old version.
Claude Opus 5 came out today. so I re-ran the whole board on it.
short version: it fixed last year's SDKs. it did not fix this year's.
The board, now on Opus 5
Same tasks, same libraries, new model:
I build a small tool called SDKProof. It measures whether AI coding agents write a library's current...
Here's a bug I kept hitting. I'd ask an AI assistant to write some code against a library — Prisma,...
Might seem like a minor release. But it really isn't.
Anthropic's new Claude Opus 4.8 aced our math problem and shipped a spotless game—then drained our entire token quota in a single…
A technical view into Anthropic's latest model.
Claude Opus 4.7 released April 16, 2026. 13% coding lift, 3.75MP vision, xhigh effort level, task budgets. Complete developer…
