Opus 5 scores 96.2% on ARC-AGI-3 with generic agent, near-human performance
inductionheads · x · 2026-08-14
Developer Jeremy Berman announced on X that using Claude Code with Opus 5 (high) achieved 96.2% accuracy on the ARC-AGI-3 benchmark, with pass@2 at 99.3%. The program is almost entirely generic, relying on agent capabilities rather than ARC-specific tweaks. Plinz commented that François Chollet's approach of challenging researchers to prove him wrong is productive and courageous.
Related event: Opus 5 Tops ARC-AGI-3 via Claude Code(7 posts)→
More from AGI Musings
- Sabine Hossenfelder fires back: academia's impact metric turned physics into a paper mill — skdh · 2026-10-16
- Researcher credits GPT for multiple accepted papers, incl. disproving Wegner's 1965 conjecture — ctjlewis · 2026-10-16
- dlwh: We optimized the legible version of our goals — and forgot the computation we meant to speed up — dlwh · 2026-10-16
- Ideas no longer have a waiting period: agents can build them as soon as they're articulable — akbirthko · 2026-10-16
- Ex-researcher: labs willing to cheat/distill ship great models faster than those building self-sufficient stacks — _aidan_clark_ · 2026-10-16
- Terry Tao reportedly still expects Riemann Hypothesis proof to be human+AI effort around 2050 — basedjensen · 2026-10-16