The Harness Effect: How Orchestration Design Sets the Token Economics of Enterprise Agentic AI

arXiv 2607.06906 ↗First posted 8 Jul 2026cs.AIcited by 1

On the map: LLM reasoning → Reasoning for agents, planning and code

also in 1 more

LLM agents and tool use → Agent memory systems

Abstract (arXiv)

Agentic AI development today runs on token maxing: buying capability with tokens -- longer reasoning traces, more turns, wider tool payloads, bigger replayed contexts -- so tokens per task grow faster than task value. Falling per-token prices mask the pattern; total spend rises anyway. We argue the decisive lever against token maxing is the harness: the orchestration layer that assembles context, exposes tools, sequences turns, delegates work, and carries enterprise observability and governance. We isolate it with a controlled swap: 22 locked evaluation tasks, six foundation models (Claude Sonnet 4.6, Gemini 3.1, Gemini Flash 3.5, Qwen 3.6, GLM 5.1, Palmyra X6), changing only the orchestration layer -- a frozen conventional production loop versus the Writer Agent Harness. Holding models constant, the harness cuts blended cost per task 41% (0.21->0.12), median wall-clock 44% (48s->27s), and tokens per task 38% (14.2k->8.8k), with task-completion quality at parity (0.78->0.81, directional at this sample size). Efficiency is model-invariant -- every model gets cheaper (33-61%) -- while quality gains are capability-dependent: a model's gain correlates almost perfectly with its baseline strength (r=0.99, n=6), a phenomenon we term harness leverage. Quality per dollar rises 82%; task-completions per million tokens rise from 54.9 to 92.0. On this workload the orchestration layer moved cost per task more than the full spread of the model menu did. We formalize token economics at the orchestration layer (including effective input price under prompt caching), detail the six mechanism families behind the effect -- cache-shape discipline to failure-spend governance -- compare six widely used agent systems on the same axes, and argue the harness is the one component whose efficiency multiplies across every model an organization runs -- present and future.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 24 of its 37 references that are in the map. First posted on arXiv 8 Jul 2026.

Builds on

24 of 37 references.

References · 24

2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
2020Scaling Laws for Neural Language Modelscited by 6,586
2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
2022Training Compute-Optimal Large Language Modelscited by 3,019
2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,988

12 more

2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,986
2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,737
2022Fast Inference from Transformers via Speculative Decodingcited by 1,653
2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
2023FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancecited by 853
2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
2025GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learningcited by 542
2025Chain of Draft: Thinking Faster by Writing Lesscited by 297
2023LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Modelscited by 287
2024Token-Budget-Aware LLM Reasoningcited by 247
2026Harness-Bench: Measuring Harness Effects across Models in Realistic Agent Workflowscited by 57
2026From Model Scaling to System Scaling: Scaling the Harness in Agentic AIcited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

2026Efficiency Matters in Autonomous Researchcited by 0

Similar, not cited

Close in wording, no citation either way