EvoRoute: Experience-Driven Self-Routing LLM Agent Systems
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917186418769920 |
|---|---|
| author | Zhang, Guibin Yu, Haiyang Yang, Kaiming Wu, Bingli Huang, Fei Li, Yongbin Yan, Shuicheng |
| author_facet | Zhang, Guibin Yu, Haiyang Yang, Kaiming Wu, Bingli Huang, Fei Li, Yongbin Yan, Shuicheng |
| contents | Complex agentic AI systems, powered by a coordinated ensemble of Large Language Models (LLMs), tool and memory modules, have demonstrated remarkable capabilities on intricate, multi-turn tasks. However, this success is shadowed by prohibitive economic costs and severe latency, exposing a critical, yet underexplored, trade-off. We formalize this challenge as the \textbf{Agent System Trilemma}: the inherent tension among achieving state-of-the-art performance, minimizing monetary cost, and ensuring rapid task completion. To dismantle this trilemma, we introduce EvoRoute, a self-evolving model routing paradigm that transcends static, pre-defined model assignments. Leveraging an ever-expanding knowledge base of prior experience, EvoRoute dynamically selects Pareto-optimal LLM backbones at each step, balancing accuracy, efficiency, and resource use, while continually refining its own selection policy through environment feedback. Experiments on challenging agentic benchmarks such as GAIA and BrowseComp+ demonstrate that EvoRoute, when integrated into off-the-shelf agentic systems, not only sustains or enhances system performance but also reduces execution cost by up to $80\%$ and latency by over $70\%$. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_02695 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | EvoRoute: Experience-Driven Self-Routing LLM Agent Systems Zhang, Guibin Yu, Haiyang Yang, Kaiming Wu, Bingli Huang, Fei Li, Yongbin Yan, Shuicheng Computation and Language Multiagent Systems Complex agentic AI systems, powered by a coordinated ensemble of Large Language Models (LLMs), tool and memory modules, have demonstrated remarkable capabilities on intricate, multi-turn tasks. However, this success is shadowed by prohibitive economic costs and severe latency, exposing a critical, yet underexplored, trade-off. We formalize this challenge as the \textbf{Agent System Trilemma}: the inherent tension among achieving state-of-the-art performance, minimizing monetary cost, and ensuring rapid task completion. To dismantle this trilemma, we introduce EvoRoute, a self-evolving model routing paradigm that transcends static, pre-defined model assignments. Leveraging an ever-expanding knowledge base of prior experience, EvoRoute dynamically selects Pareto-optimal LLM backbones at each step, balancing accuracy, efficiency, and resource use, while continually refining its own selection policy through environment feedback. Experiments on challenging agentic benchmarks such as GAIA and BrowseComp+ demonstrate that EvoRoute, when integrated into off-the-shelf agentic systems, not only sustains or enhances system performance but also reduces execution cost by up to $80\%$ and latency by over $70\%$. |
| title | EvoRoute: Experience-Driven Self-Routing LLM Agent Systems |
| topic | Computation and Language Multiagent Systems |
| url | https://arxiv.org/abs/2601.02695 |