Introducing Muse Spark Open, The Most Capable US Open-Weight Model to Date

Benchmark Model Rank Results
code-generation-on-scicodeMuse Spark Open#2Accuracy: 58.5
visual-question-answering-vqa-on-simplevqaMuse Spark Open#2Accuracy: 73.6
agents-on-jobbench-main-65-tasksMuse Spark Open#4Score: 61.8
multiple-choice-question-answering-mcqa-on-gpqa-diamondMuse Spark Open#16Accuracy: 89.0
code-generation-on-terminal-bench-2-1Muse Spark Open (Terminus 2 agent)#21Accuracy: 76.4