AI RESEARCH

RADAR: Reasoning-Ability and Difficulty-Aware Routing for Reasoning LLMs

arXiv CS.AI

ArXi:2509.25426v3 Announce Type: replace Reasoning language models have nstrated remarkable performance on many challenging tasks in math, science, and coding. Choosing the right reasoning model for practical deployment involves a performance and cost tradeoff at two key levels: model size and reasoning budget, where larger models and higher reasoning budget lead to better performance but with increased cost and latency.