Agentic Performance at the Edge: Insights from Benchmarking
On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks
Abstract (arXiv)
Agentic artificial intelligence (AI) is a natural fit for Internet of Things (IoT) and edge systems, but edge deployments are often constrained to models around 8 billion parameters or smaller. An important question is: How much agentic-task quality is lost when model size is constrained by memory, power, and latency budgets? To address this question, in this paper, we provide an initial empirical study considering edge-focused model scaling, general-purpose versus coder-oriented model effects, and tool-enabled execution under a fixed protocol. We introduce a domain-conditioned evaluation methodology, an implementation-grounded analysis of model-tool interactions, practical guidance for model selection under constraints, and an analysis of failure modes that reveals distinct semantic versus execution failure patterns across model families. Our core finding is that edge-agent quality is not a simple function of parameter count. Robust deployment depends on the joint design of model choice and tool workflow. Domain-conditioned analysis reveals Pareto fronts in the accuracy-latency space that can guide strategy selection based on operational priorities.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 14 of its 19 references that are in the map. First posted on arXiv 11 May 2026.
Builds on
14 of 19 references.
Closest ancestors · 2
Same direction, newest first
- 2024OfficeBench: Benchmarking Language Agents across Multiple Applications for Office Automation26 Jul
- 2023AgentBench: Evaluating LLMs as Agents7 Aug
Other references · 12
- 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
- 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
- 2022Large Language Models are Zero-Shot Reasonerscited by 4,374
- 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
- 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,418
- 2024Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phonecited by 1,917
- 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
- 2024TinyLlama: An Open-Source Small Language Modelcited by 618
- 2024RouterDC: Query-Based Router by Dual Contrastive Learning for Assembling Large Language Modelscited by 88
- 2025ITBench: Evaluating AI Agents across Diverse Real-World IT Automation Taskscited by 51
- 2025Device-Cloud Collaborative LLM Inference with Multi-Modal, Multi-Task, Multi-Turn Conversationscited by 14
- 2025MESS+: Dynamically Learned Inference-Time LLM Routing in Model Zoos with Service Level Guaranteescited by 9
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026A Self-Calibrating Agentic AI Framework for Autonomous Edge Resource Allocation
- 2025AgentArch: A Comprehensive Benchmark to Evaluate Agent Architectures in Enterprise
- 2025Beyond Black-Box Benchmarking: Observability, Analytics, and Optimization of Agentic Systems
- 2026Where Should Agents Live? Energy-Memory Characterization of Agentic AI for the Edge-Cloud Continuum
- 2026AI-NativeBench: An Open-Source White-Box Agentic Benchmark Suite for AI-Native Systems
- 2025Edge General Intelligence Through World Models and Agentic AI: Fundamentals, Solutions, and Challenges
- 2025Towards a Science of Scaling Agent Systems
- 2025Beyond Accuracy: A Multi-Dimensional Framework for Evaluating Enterprise Agentic AI Systems