Frontier lab rumor says Opus 5 ARC-AGI 3 score looks fake
flowersslop · x · 2026-07-25
A person claiming to be inside a frontier AI lab says the reported Opus 5 ARC-AGI 3 score “looks fake,” suggesting either aggressive benchmark gaming or a serious underestimation by the competing lab.
The post does not provide proof, but it captures a growing rumor that the result may be less about genuine capability and more about benchmark manipulation.
Related event: Anthropic Accused of Gaming ARC-AGI-3(2 posts)→
More from Models
- Opus 5 chart shows competitive gains across coding, search and computer use — CtrlAltDwayne · 2026-07-25
- Live SlopCodeBench run puts Opus 5 ahead of Opus 4.8 and Sonnet 5 — HamelHusain · 2026-07-25
- Hands-on: A New Cost-Effective and Fast Model for Agent Workflows — doodlestein · 2026-07-25
- Claude Opus 5 reportedly beats Fable 5 on a hard 3D coding test at 75% of the price — rohanpaul_ai · 2026-07-25
- Chart says Claude Opus 5 blocks far less defensive coding than Fable 5 — repligate · 2026-07-25
- Anthropic Launches Opus 5: Focuses on Token Efficiency Over Capability Leaps — Ars Technica AI · 2026-07-25