Newest papers

RouterWorld: Scaling Conversational AI with Latency-Aware LLM Routing

Jinyang Li, Denis Emelin, Simone Tedeschi, Weiwei Cheng, Cristina Morariu, Haytham Assem, Jens Lehmann, Emine Yılmaz, Angus Addlesee · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26511 · cited by 0 works (OpenAlex) · open-access PDF

Portuguese

Os sistemas de código de AI de conversação construídos com base em Modelos de Dados de Grandes Línguas (LLMs) requerem mecanismos de roteamento que equilibram a qualidade de resposta e a latência de resposta em escala. No entanto, os benchmarks de roteamento LLM existentes assumem a plena observabilidade de todos os resultados do modelo de candidato, sub-enfatizam a latência como uma métrica de primeira classe e a avaliação de fragmentos em protocolos heterogêneos desalinhados com configurações do mundo real. Introduzimos o RouterWorld, um benchmark abrangente para o roteamento de LLM em GenAI de conversação. Crucialmente, o RouterWorld unifica as respostas de um único turno de 371k e as trajetórias de agentes de vários turnos de 76k, adota

Machine translation (Meta M2M100); the English original follows.

Abstract

Conversational AI systems built on Large Language Models (LLMs) require routing mechanisms that balance answer quality and response latency at scale. However, existing LLM routing benchmarks assume full observability of all candidate model outcomes, under-emphasize latency as a first-class metric, and fragment evaluation across heterogeneous protocols misaligned with real-world settings. We introduce RouterWorld, a comprehensive benchmark for LLM routing in conversational GenAI. RouterWorld unifies 371k single-turn responses and 76k multi-turn agentic trajectories, adopts natural user interactions with a standardized LLM-as-a-judge protocol, and decouples efficiency from transient pricing via hardware-grounded latency annotations. Crucially, RouterWorld formalizes routing under partial observability through two setups: Cold-Start, which restricts learning to sparse, static logs, and Voyager, which models continuous data streams where routing decisions shape future training distributions. Evaluating 13 routing baselines, we find that most routers over-emphasize accuracy while ignoring latency under partial observability. Code and benchmark data are publicly available.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Observability, Latency (audio), Static routing, Benchmark (surveying), Link-state routing protocol

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange