Meta · LIVEBENCH 2026-06-25

Muse Spark 1.2 xHigh Effort

Published benchmark results for this specific model configuration.

muse-spark-1.2-xhigh

Overall score78.0 / 100

Cost / successful task$0.375USD · benchmark workload

Weight accessNot reported

Where this configuration performs

Category averages from the same release, on a 0–100 scale.

Reasoning90.0

Coding77.5

Agentic Coding57.6

Mathematics91.2

Data Analysis76.5

Language78.6

Instruction Following74.3

Task-level results

theory of mind80.8zebra puzzle95.3spatial96.0logic with navigation88.0code generation74.6code completion80.4javascript72.7typescript50.0python50.0AMPS Hard89.0integrals with game91.0math comp97.1olympiad87.8consecutive events82.5tablejoin46.8tablereformat100.0connections100.0plot unscrambling61.7typos74.0paraphrase74.3simplify68.7story generation73.7summarize80.6

Source files: Scores ↗ Categories ↗ Costs ↗ . Missing cost is unknown, never free. Compare configurations within the same release.