Muse Spark 1.1 | Terminal-Bench 2.1 — mini-swe-agent

Meta AI (originally Facebook AI Research)'s Muse Spark 1.1 evaluation results on Terminal-Bench 2.1 — mini-swe-agent

Score: 80.0

Reasoning level: xhigh

Source

Muse Spark 1.1 (Registry No. 80002)

Meta AI (originally Facebook AI Research)

Terminal-Bench 2.1 — mini-swe-agent