# Are Tools All We Need? Unveiling the Tool-Use Tax in LLM Agents

arXiv: https://arxiv.org/abs/2605.00136 · HTML: https://mlsanity.com/paper/2605.00136/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Placed in Tool learning and function calling, one of 16 directions in LLM agents and tool use.
- Builds on 18 of its 35 references that are in the map.
- First posted on arXiv 30 Apr 2026.

## Placed in

- LLM agents and tool use / [Tool learning and function calling](https://mlsanity.com/agents/tool-learning-and-function-calling/)

## Builds on (most cited first, 18 of 18)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs](https://mlsanity.com/paper/2307.16789/) (2023)
- [CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing](https://mlsanity.com/paper/2305.11738/) (2023)
- [ToolACE: Winning the Points of LLM Function Calling](https://mlsanity.com/paper/2409.00920/) (2024)
- [SealQA: Raising the Bar for Reasoning in Search-Augmented Language Models](https://mlsanity.com/paper/2506.01062/) (2025)
- [How Is LLM Reasoning Distracted by Irrelevant Context? An Analysis Using a Controlled Benchmark](https://mlsanity.com/paper/2505.18761/) (2025)
- [Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents](https://mlsanity.com/paper/2510.02837/) (2025)
- [AgentNoiseBench: Benchmarking Robustness of Tool-Using LLM Agents Under Noisy Condition](https://mlsanity.com/paper/2602.11348/) (2026)
- [Dissecting Logical Reasoning in LLMs: A Fine-Grained Evaluation and Supervision Study](https://mlsanity.com/paper/2506.04810/) (2025)
- [PaperArena: An Evaluation Benchmark for Tool-Augmented Agentic Reasoning on Scientific Literature](https://mlsanity.com/paper/2510.10909/) (2025)
- [Can LLMs Correct Themselves? A Benchmark of Self-Correction in LLMs](https://mlsanity.com/paper/2510.16062/) (2025)
- [Demystifying the Lifecycle of Failures in Platform-Orchestrated Agentic Workflows](https://mlsanity.com/paper/2509.23735/) (2025)
- [TTPA: Token-level Tool-use Preference Alignment Training Framework with Fine-grained Evaluation](https://mlsanity.com/paper/2505.20016/) (2025)
- [Lost in the Noise: How Reasoning Models Fail with Contextual Distractors](https://mlsanity.com/paper/2601.07226/) (2026)
- [Investigating Tool-Memory Conflicts in Tool-Augmented LLMs](https://mlsanity.com/paper/2601.09760/) (2026)
- [A Survey on Evaluating Quality and Trustworthiness in LLM-Generated Data](https://mlsanity.com/paper/2601.17717/) (2026)
- [Assessing LLM Reasoning Steps via Principal Knowledge Grounding](https://mlsanity.com/paper/2511.00879/) (2025)

## Built on by (most cited first, 11 of 11)

- [When the Tool Decides: LLM Agents Defer Blindly to Graph Neural Network Tools, and Stronger Backbones Defer More](https://mlsanity.com/paper/2606.14476/) (2026)
- [Do Multimodal Agents Really Benefit from Tool Use? A Systematic Study of Capability Gains](https://mlsanity.com/paper/2606.02357/) (2026)
- [ToolGate: Token-Efficient Pre-Call Control for Tool-Augmented Vision-Language Agents](https://mlsanity.com/paper/2606.03054/) (2026)
- [Reasoning effort, not tool access, buys first-try reliability in agentic code generation: an observational study](https://mlsanity.com/paper/2607.02436/) (2026)
- [Reinforcement Learning for Large Language Model Selective Evidence Adoption from Contaminated Retrieval Results](https://mlsanity.com/paper/2607.20090/) (2026)
- [When Does Restricting a Coding Agent to executecode Help? A Regime × Agent-Design Ablation](https://mlsanity.com/paper/2607.10569/) (2026)
- [EMR: Self-Evolving Medical Multi-Agent System via Experience Mining and Reuse](https://mlsanity.com/paper/2609.15161/) (2026)
- [PHREEQC-MCQ-200: A Diagnostic Benchmark for Tool-Augmented Scientific Simulator Agents](https://mlsanity.com/paper/2607.00436/) (2026)
- [Evaluating LLMs in Database Scenarios: A Lifecycle Benchmark for Assessing Their Potential in Core Database Tasks](https://mlsanity.com/paper/2608.03794/) (2026)
- [What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics](https://mlsanity.com/paper/2608.16370/) (2026)
- [AsynCodeBench: Benchmarking Collaboration of Asynchronous Multi-Agent Systems in Software Engineering](https://mlsanity.com/paper/2609.32662/) (2026)

## Abstract (arXiv)

Tool-augmented reasoning has become a popular direction for LLM-based agents, and it is widely assumed to improve reasoning and reliability. However, we demonstrate that this consensus does not always hold: in the presence of semantic distractors, tool-augmented reasoning does not necessarily outperform native CoT. To explain this performance gap, we propose a Factorized Intervention Framework that isolates the cost of prompt formatting, the overhead of the tool-calling protocol, and the actual gain from executing tools. Our analysis reveals a critical tradeoff: under semantic noise, the gains from tools often fail to offset the "tool-use tax", which is the performance degradation introduced by the tool-calling protocol itself. To address this, we introduce G-STEP, a lightweight inference-time gate to mitigate protocol-induced errors. While this yields partial recovery, our findings suggest that more substantial improvements still require strengthening the model's intrinsic reasoning and tool-interaction capabilities.