Muse Spark 1.3（max）在 Tau3-Bench Banking 上得分 52%，成为该评测新的第一名。Muse Spark 1.3（xhigh）得分 47%，与 Claude Fable 5.1（max，47%）和 GLM-5.3-Flash（47%）持平，落后于其同门版本，也落后于 Qwen3.8 Max（51%）、Grok 4.6（high，51%）和 GLM-5.3（max，50%）等。相比上月的 Muse Spark 1.2（35%），该评测上的显著提升是各 Muse Spark 1.3 版本 Artificial Analysis Intelligence Index 得分提高的最大单一驱动因素
Muse Spark 1.3 (max) scores 52% on Tau3-Bench Banking, the new #1 performer on this evaluation. Muse Spark 1.3 (xhigh) scores 47%, level with Claude Fable 5.1 (max, 47%) and GLM-5.3-Flash (47%) and behind its sibling in addition to e.g. Qwen3.8 Max (51%), Grok 4.6 (high, 51%), and GLM-5.3 (max, 50%). The significant gains in this evaluation vs. Muse Spark 1.2 last month (35%) are the largest single driver of each Muse Spark 1.3 variant’s improved Artificial Analysis Intelligence Index score
来源：@ArtificialAnlys · x.com