method · first seen 6 Nov 2026
per-query cost prediction for LLM routing
Also: cost-aware LLM routing
Predicting each query's output token cost for each candidate model so routers can choose models by expected cost and accuracy.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Scout Before You Route: Predicting Task Success and Token Count from a Small LLM PrefillPapers · 6 Nov 2026 · Aristides Milios, Christopher Pal
- When the Cheapest Feasible Router Fails: Auditing Deployment Commitments in Cost-Aware LLM RoutingPapers · 16 Nov 2026 · Minxuan Hu, Qishi Zhan, Ziheng Chen, Zhengtao Yao, Zihan Dong, Jiayu Yi, Wenxi S…
Latest sources
- When the Cheapest Feasible Router Fails: Auditing Deployment Commitments in Cost-Aware LLM RoutingPapers · 16 Nov 2026 · Minxuan Hu, Qishi Zhan, Ziheng Chen, Zhengtao Yao, Zihan Dong, Jiayu Yi, Wenxi S…
- Scout Before You Route: Predicting Task Success and Token Count from a Small LLM PrefillPapers · 6 Nov 2026 · Aristides Milios, Christopher Pal