Muse Spark 1.3 (Max): benchmark results

Provider: Meta. Released 2026-09-02. Access: API.

Unified ELO 1743 ± 1, rank #39 of 1761 rated models, from 18 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Tau3 Banking52.37Success Rate (%)100
AA Long Context Reasoning84.33Accuracy (%)99.2
AA GDPval1720.15ELO98.7
Artificial Analysis Intelligence Index52.95Intelligence Index98.7
AA GPQA Diamond93.84Accuracy (%)98.4
AA Humanity's Last Exam49.07Accuracy (%)97.5
AA CritPt24.86Accuracy (%)94.8
AA Omniscience24.93Score94.2
AA SciCode58.33Accuracy (%)90.8
AA Omniscience - Business38.3Accuracy (%)89.4
AA Omniscience - Science, Engineering & Mathematics46.9Accuracy (%)89.4
AA Omniscience - Law38.3Accuracy (%)87.7

Interactive version: theaggregate.ai/model?slug=muse-spark-1-3-max · How It Works · Data refreshed daily, snapshot 2026-09-05.