HybridFlow: Resource-Adaptive Subtask Routing for Efficient Edge-Cloud LLM Inference

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dong, Jiangwen, Li, Jiayu, Zheng, Tianhang, Lin, Wanyu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918312650211328
author Dong, Jiangwen
Li, Jiayu
Zheng, Tianhang
Lin, Wanyu
author_facet Dong, Jiangwen
Li, Jiayu
Zheng, Tianhang
Lin, Wanyu
contents Edge-cloud collaborative inference is becoming a practical necessity for LLM-powered edge devices: on-device models often cannot afford the required reasoning capability, while cloud-only inference could be prohibitively costly and slow under strict latency and token/API budgets. However, existing edge-cloud collaboration methods often route per query or fixed steps simply based-on the estimated difficulty. Such coarse and static heuristics overlook subtask dependencies, missing opportunities for parallel execution and budget-adaptive routing. To this end, we propose \textbf{HybridFlow}, a resource-adaptive edge-cloud inference framework that (i) builds a dependency-aware DAG for each query and executes newly unlocked subtasks in parallel, reducing end-to-end latency; (ii) routes each subtask online to the edge or cloud via a learned benefit--cost utility model that dynamically trades accuracy gains against token/API and latency budgets, thereby reducing unnecessary cloud usage while preserving reasoning quality. Across GPQA, MMLU-Pro, AIME24, and LiveBench-Reasoning, HybridFlow improves the cost-accuracy trade-off, reducing latency and cloud API usage while maintaining competitive accuracy against strong structured reasoning baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2512_22137
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HybridFlow: Resource-Adaptive Subtask Routing for Efficient Edge-Cloud LLM Inference
Dong, Jiangwen
Li, Jiayu
Zheng, Tianhang
Lin, Wanyu
Distributed, Parallel, and Cluster Computing
Edge-cloud collaborative inference is becoming a practical necessity for LLM-powered edge devices: on-device models often cannot afford the required reasoning capability, while cloud-only inference could be prohibitively costly and slow under strict latency and token/API budgets. However, existing edge-cloud collaboration methods often route per query or fixed steps simply based-on the estimated difficulty. Such coarse and static heuristics overlook subtask dependencies, missing opportunities for parallel execution and budget-adaptive routing. To this end, we propose \textbf{HybridFlow}, a resource-adaptive edge-cloud inference framework that (i) builds a dependency-aware DAG for each query and executes newly unlocked subtasks in parallel, reducing end-to-end latency; (ii) routes each subtask online to the edge or cloud via a learned benefit--cost utility model that dynamically trades accuracy gains against token/API and latency budgets, thereby reducing unnecessary cloud usage while preserving reasoning quality. Across GPQA, MMLU-Pro, AIME24, and LiveBench-Reasoning, HybridFlow improves the cost-accuracy trade-off, reducing latency and cloud API usage while maintaining competitive accuracy against strong structured reasoning baselines.
title HybridFlow: Resource-Adaptive Subtask Routing for Efficient Edge-Cloud LLM Inference
topic Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2512.22137