LLM-based agents deliver state-of-the-art performance across tasks but incur high end-to-end latency on edge devices. We introduce Agent-X, a software-only, accuracy-preserving framework that accelerates both the prefill and decode stages of on-device agent workloads. Agent-X's two key components rewrite prompts to leverage prefix caching tailored to agent-specific input-token patterns and enable LLM-free speculative decoding for fast token generation with minimal overhead. On representative agentic workloads, Agent-X achieves a 1.61x end-to-end speedup in real systems with no accuracy loss and can be seamlessly integrated into existing on-device AI agents. To the best of our knowledge, ours is the first to systematically characterize and eliminate latency bottlenecks in on-device agents.
@article{arxiv.2605.10380,
title = {Agent-X: Full Pipeline Acceleration of On-device AI Agents},
author = {Jinha Chung and Byeongjun Shin and Jiin Kim and Minsoo Rhu},
journal= {arXiv preprint arXiv:2605.10380},
year = {2026}
}