Which Model for Which Role? Cost-Efficient Pareto Recommendation for LLM Agentic Workflows
Abstract
LLM agentic workflows increasingly employ different models for specialized roles. Users may prefer different trade-offs between task quality and deployment cost, sometimes also considering latency. Finding good assignments incurs a separate search cost from evaluating multiple candidates. We introduce Cost-Coupled Gittins (CC-Gittins), a multi-objective extension of single-objective cost-aware Gittins search for Pareto recommendation. It exploits the coupling between evaluation and deployment costs: assignments that cost more to deploy also cost more to evaluate. The search alternates between seeking higher quality and lower deployment cost, reusing each evaluation to inform both searches. After each batch, CC-Gittins recommends assignments that are nondominated under uncertainty-adjusted estimates of quality and deployment cost. We evaluate CC-Gittins on eight quality--cost matrices from workflows with two to five roles and up to 1,681 candidate assignments. We also evaluate the method with latency as a third objective. Across the eight benchmarks, CC-Gittins recovers a substantial fraction of the true Pareto frontier at roughly 10% of the exhaustive-evaluation cost, and recovers most of the frontier with few false positives by 30%.
Related
- Principled Human-AI Collaboration in Agentic Workflows
- New Pareto Z-eigenvalue localization theorems for tensors with applications
- Pareto optimality under nondegeneracy at infinity
- Convergence-Aware Pareto Selection of Covariate Scaling Transformations for Markov Deterioration Hazard Models: Evidence from Bridge Inspection Data
- Full asymptotics for the parabolic Anderson model with Pareto potential in the heaviest-tailed regime
BibTeX
@misc{xie2026which,
title = {{Which Model for Which Role? Cost-Efficient Pareto Recommendation for LLM Agentic Workflows}},
author = {Qian Xie and Wenyue Hua and Yueli He and Sripad Karne and Armaan Agrawal and Nikos Pagonas and Eugene Wu and Kostis Kaffes and Tianyi Peng and Nairen Cao},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.26676},
eprint = {2610.26676},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.26676},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Which Model for Which Role? Cost-Efficient Pareto Recommendation for LLM Agentic Workflows
AU - Qian Xie
AU - Wenyue Hua
AU - Yueli He
AU - Sripad Karne
AU - Armaan Agrawal
AU - Nikos Pagonas
AU - Eugene Wu
AU - Kostis Kaffes
AU - Tianyi Peng
AU - Nairen Cao
PY - 2026
DA - 2026/10/29
PB - arXiv
DO - 10.48550/arxiv.2610.26676
UR - https://arxiv.org/abs/2610.26676
AB - LLM agentic workflows increasingly employ different models for specialized roles. Users may prefer different trade-offs between task quality and deployment cost, sometimes also considering latency. Finding good assignments incurs a separate search cost from evaluating multiple candidates. We introduce Cost-Coupled Gittins (CC-Gittins), a multi-objective extension of single-objective cost-aware Gittins search for Pareto recommendation. It exploits the coupling between evaluation and deployment costs: assignments that cost more to deploy also cost more to evaluate. The search alternates between seeking higher quality and lower deployment cost, reusing each evaluation to inform both searches. After each batch, CC-Gittins recommends assignments that are nondominated under uncertainty-adjusted estimates of quality and deployment cost. We evaluate CC-Gittins on eight quality--cost matrices from workflows with two to five roles and up to 1,681 candidate assignments. We also evaluate the method with latency as a third objective. Across the eight benchmarks, CC-Gittins recovers a substantial fraction of the true Pareto frontier at roughly 10% of the exhaustive-evaluation cost, and recovers most of the frontier with few false positives by 30%.
ER -