Meta AI Scores Perfect 30/30 on Physics Olympiad Exam

Meta AI's model reportedly scored a perfect 30 out of 30 on the Asian Physics Olympiad's theoretical exam. This achievement, if verified, marks a significant milestone in AI reasoning. What's interesting is that the gold medal threshold for this Olympiad typically ranges from 21 to 23 out of 30. So, a perfect score doesn't just clear the bar; it vaults over it by a comfortable margin. Meta's 'Muse Spark' model has been leading on multiple 2025 International Physics Olympiad leaderboards. This positions Meta as a serious contender in scientific reasoning benchmarks. Other players like Google's Gemini agents and Shanghai AI Lab's P1 model have also achieved near-perfect theoretical scores in controlled tests. The HiPhO benchmark suite, published in September 2025, evaluates AI performance across 12 different physics olympiads. While top models perform strongly, they have generally fallen short of perfection, clustering around gold-medal thresholds. The bottom line is that this potential achievement signals a new frontier in AI reasoning capabilities, with implications for how tech giants compete in the artificial intelligence arms race.