Meta Muse Wonder 跑分反超 GPT-6 Astra
Meta 新模型 Muse Wonder 在个人超级智能指数上拿 59.1，高于 GPT-6 Astra 的 54.5 和 Opus 5.5 的 55.2，但整体指数仍落后 Opus。
Meta is back in the frontier model race! Muse Wonder is starting to appear, and the first benchmark numbers look seriously impressive. Muse Wonder Full scores 59.1 on the Personal Superintelligence Index, beating GPT-6 Astra (54.5) and Claude Opus 5.5 (55.2). It also leads on several agentic benchmarks, including legal work, business automation, and long-horizon coding. The coding results are particularly interesting: 79.0 on DeepSWE 1.1 and 57.5 on SWE-Marathon, comfortably ahead of Opus 5.5. However, Opus still has a substantial lead on ProgramBench and remains ahead on the overall Artificial Analysis Index. Meta is becoming a serious contender again, especially when it comes to agentic capabilities. And competition at the frontier is exactly what we want!