cs.DCNov 12, 2026

Reducing Batch Makespan for LLM Agent Workloads through Inference-Environment Coordination

Authors: Wei Feng, Meiling Chen, Boxiao Chen, Lu Liu, Bin Zi, Libin Ye, Xingyu Lu, Lei Xu, 

Organizations: Ling Team, Ant Group

Abstract

Efficient batch execution of agent tasks is critical to LLM training and development workflows, including benchmark evaluation, synthetic-data generation, and RL rollouts. Since these tasks alternate between LLM inference and environment interaction, we present coordinated optimizations to reduce batch makespan, the time required to complete all tasks in a batch. For hybrid LLMs, a workload-based resource model guides GPU-memory allocation between key-value (KV) cache and recurrent state and provides a reference sample concurrency to exploit parallelism while limiting recomputation. Trajectory-aware routing and scheduling promote cache reuse while balancing worker load. Under constrained environment capacity, we decouple sample admission from environment execution and introduce Chain-Advancing Action Scheduling (CAAS), which prioritizes long chains using observed progress to shorten the batch tail. Controlled replay combines actual inference with recorded environment responses and durations for ablations of individual optimizations. Live evaluations with Ling-3.0-Flash and Ling-3.1-Flash on the BFCL, SWE-bench Multilingual, WideSearch, and MCP-Atlas benchmarks demonstrate makespan reductions of 10.4% to 65.0%, while evaluation scores remain aligned. Furthermore, these methods have been successfully applied to training and evaluation workflows for the Ling-3.x model series, accelerating the model-development cycle.

Figures & tables

Appendix figures & tables3 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Pythia: Exploiting Workflow Predictability for Efficient Agent-Native LLM Serving

    Apr 28, 2026Shan Yu, Junyi Shu, Yuanjiang Ni +14LLM Inference EfficiencyMulti-Agent LLM Systems

  2. Threshold-Based Exclusive Batching for LLM Inference

    May 30, 2026Weifang Zhang, Yuzhou Nie, Bowen Pang +2GPU AccelerationLLM Inference

  3. LLM4LLM: Bridging Kernel Benchmarks and Real Deployment via Closed-Loop Agentic Optimization

    Aug 22, 2026Hui Zeng, Pengfei Yang, Yanxin Chen +2LLM Inference AccelerationGPU Kernel Optimization