OneBench
Dynamic Model Routing and Cascading for Efficient LLM Inference: A Survey | OneBench: AI Insights