Do Multimodal Agents Really Benefit from Tool Use? A Systematic Study of Capability Gains

• URL: http://arxiv.org/abs/2606.02357v1
• Date: Mon, 01 Jun 2026 15:04:25 GMT
• Title: Do Multimodal Agents Really Benefit from Tool Use? A Systematic Study of Capability Gains
• Abstract summary: We study two representative thinking with images'' agents, Thyme and DeepEyesV2, across real-world understanding, OCR, chart understanding, and mathematical reasoning.<n>We find that tool access yields little consistent aggregate improvement, does not reliably reduce generated-token cost, and leaves only a small tool-only solved set.
• Score: 20.11899219087138
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Tool-augmented multimodal agents show strong benchmark gains, often taken as evidence that agents have learned to use tools. We argue that this interpretation can be premature: a tool-call trace alone does not show whether the tool supplied answer-critical information. We study two representative ``thinking with images'' agents, Thyme and DeepEyesV2, across real-world understanding, OCR, chart understanding, and mathematical reasoning. Each agent is compared with its Tool-Free counterpart and with a Pure-Text Reasoner trained from the same source pool without tool-calling trajectories. Tool access yields little consistent aggregate improvement, does not reliably reduce generated-token cost, and leaves only a small tool-only solved set: 93% of DeepEyesV2's tool-solved problems and 96% of Thyme's are also solved by at least one non-tool setting. Mechanism ablations further show that the full tool-use loop does not consistently outperform either the tool-call format or the returned execution result alone. In the settings we study, the analyzed agents appear to learn tool-calling patterns more reliably than tool-contributed capabilities, suggesting that evaluation should distinguish tool availability from whether tools actually expand what agents can solve.

Related papers

• Agents Trust Tools Too Much: Measuring Reliance on Unreliable Tools [16.535602696436424]
  We evaluate how agents respond to unreliable tool returns using three tools-web search, LLM sub-agent delegation, and code execution.<n>For each tool, we corrupt its returns and measure whether agents adopt the corrupted content in their final answers.<n>Agents exhibit high levels of overtrust across all three settings.
  arXiv  Detail & Related papers  (2026-09-04T17:02:32Z)
• Forgotten in Weights, Recovered by Tools: Agentic Tool Unlearning for LLM Agents [54.050506620820784]
  Agentic Tool Unlearning (ATU) aims to reduce both parametric recall and tool-mediated recovery while preserving normal tool use for retained knowledge.<n>ATU achieves a better balance between target forgetting and retained utility, making unlearning more robust under tool-augmented agent deployment.
  arXiv  Detail & Related papers  (2026-08-21T18:34:48Z)
• ToolAtlas: Learning Once, Reusing Everywhere with Tool-Side Memory [49.08868447320197]
  Large language model (LLM) agents increasingly rely on external tools served by shared providers.<n>Existing approaches improve tool use on the agent side through parameter updates, prompt refinement, or agent-side memory.<n>We introduce ToolAtlas, a graph-based framework that builds a persistent provider-side tool memory of tool capabilities.
  arXiv  Detail & Related papers  (2026-07-13T06:01:41Z)
• ToolFailBench: Diagnosing Tool-Use Failures in LLM Agents [0.0]
  ToolFailBench is a diagnostic benchmark for measuring tool-use failures.<n>We label each trace with Tool-Skip, Result-Ignore, Output-Fabrication, and Unnecessary-Tool-Use.
  arXiv  Detail & Related papers  (2026-07-06T05:25:21Z)
• SING: Synthetic Intention Graph for Scalable Active Tool Discovery in LLM Agents [64.59100414726556]
  Large language model (LLM) agents increasingly rely on harnesses that manage context, tools, and multi-turn execution.<n>Retrieval-augmented tool selection offers a natural alternative, but existing one-shot retrieval methods fail to align isolated tool descriptions with the agent's true task intention.<n>We propose SING, an intention-aware active tool discovery framework that builds an intention-tool graph linking user intentions, tool capabilities, and tool collaboration patterns.
  arXiv  Detail & Related papers  (2026-06-15T11:37:37Z)
• TRACER: Verifiable Generative Provenance for Multimodal Tool-Using Agents [19.156453695628013]
  We introduce TRACER, a framework for verifiable generative provenance in multimodal tool-using agents.<n>TraCER generates each answer sentence together with a structured provenance record that identifies the supporting tool turn, evidence unit, and semantic support relation.<n>With Qwen3-VL-8B, TRACER reaches 78.23% answer accuracy and 95.72% summary accuracy, outperforming the strongest closed-source tool-augmented baseline by 23.80 percentage points.
  arXiv  Detail & Related papers  (2026-05-11T03:32:55Z)
• Open, Reliable, and Collective: A Community-Driven Framework for Tool-Using AI Agents [22.64138018985385]
  We argue that failures stem from both tool-use accuracy (how well an agent invokes a tool) and intrinsic tool accuracy (the tool's own correctness)<n>We introduce OpenTools, a community-driven toolbox that standardizes tool schemas.<n>OpenTools includes the core framework, an initial tool set, evaluation pipelines, and a contribution protocol.
  arXiv  Detail & Related papers  (2026-03-31T18:42:36Z)
• ToolMATH: A Math Tool Benchmark for Realistic Long-Horizon Multi-Tool Reasoning [11.99927786717109]
  ToolMATH turns math problems into a controlled, correctness-checkable benchmark with tool sets.<n>ToolMATH provides actionable diagnostic evidence of failure modes in tool-augmented agents.
  arXiv  Detail & Related papers  (2026-02-24T09:23:12Z)
• ToolLibGen: Scalable Automatic Tool Creation and Aggregation for LLM Reasoning [80.10274552177096]
  Large Language Models (LLMs) equipped with external tools have demonstrated enhanced performance on complex reasoning tasks.<n>The widespread adoption of this tool-augmented reasoning is hindered by the scarcity of domain-specific tools.<n>We propose a systematic approach to automatically an unstructured collection of tools into a structured tool library.
  arXiv  Detail & Related papers  (2025-10-09T04:11:16Z)
• Tool Graph Retriever: Exploring Dependency Graph-based Tool Retrieval for Large Language Models [43.50789219459378]
  We propose Tool Graph Retriever (TGR), which exploits the dependencies among tools to learn better tool representations for retrieval.<n>First, we construct a dataset termed TDI300K to train a discriminator for identifying tool dependencies.<n>Then, we represent all candidate tools as a tool dependency graph and use graph convolution to integrate the dependencies into their representations.
  arXiv  Detail & Related papers  (2025-08-07T08:36:26Z)
• Acting Less is Reasoning More! Teaching Model to Act Efficiently [87.28134636548705]
  Tool-integrated reasoning augments large language models with the ability to invoke external tools to solve tasks.<n>Current approaches typically optimize only for final correctness without considering the efficiency or necessity of external tool use.<n>We propose a framework that encourages models to produce accurate answers with minimal tool calls.<n>Our approach reduces tool calls by up to 68.3% and improves tool productivity by up to 215.4%, while maintaining comparable answer accuracy.
  arXiv  Detail & Related papers  (2025-04-21T05:40:05Z)
• Re-Invoke: Tool Invocation Rewriting for Zero-Shot Tool Retrieval [47.81307125613145]
  Re-Invoke is an unsupervised tool retrieval method designed to scale effectively to large toolsets without training. We employ a novel multi-view similarity ranking strategy based on intents to pinpoint the most relevant tools for each query. Our evaluation demonstrates that Re-Invoke significantly outperforms state-of-the-art alternatives in both single-tool and multi-tool scenarios.
  arXiv  Detail & Related papers  (2024-08-03T22:49:27Z)
• EASYTOOL: Enhancing LLM-based Agents with Concise Tool Instruction [56.02100384015907]
  EasyTool is a framework transforming diverse and lengthy tool documentation into a unified and concise tool instruction. It can significantly reduce token consumption and improve the performance of tool utilization in real-world scenarios.
  arXiv  Detail & Related papers  (2024-01-11T15:45:11Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.