Fable 5.1 is still beating Astra on the benchmarks.
Aggregate intelligence index: Fable ahead. Terminal Bench: Fable 91, Astra 88. Finance and accounting: Astra behind even Opus 5. None of that made the launch write-ups. A 99.9% headline is easier to publish than a table.
Opus 5 spent 2h10m and ~$42 (57.8M tokens, 3 submissions, 2 rejections) on a bug where some spruces rendered flat and brown — and didn't fix it.
Gave the writeup to GPT-6 Astra, asked only for root cause + plan: zero normals past the first shadow cascade. Handed the plan back to
We built four fortnite concept skins of ai ceos for $1.39 in image calls
Elonmusk, finkd, JensenHuang and DarioAmodei, rebuilt as our thehype robot. no game engine, no 3d artist, nothing modeled by hand. the whole build ran in one claude code session on opus 5 (medium).
You have reached the end of the archive
All of Claude Opus 5