# MM-ToolSandBox: A Unified Framework for Evaluating Visual Tool-Calling Agents

arXiv: https://arxiv.org/abs/2607.11818 · HTML: https://mlsanity.com/paper/2607.11818/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal and embodied agents, one of 16 directions in LLM agents and tool use.
- Builds on 32 of its 69 references that are in the map.
- First posted on arXiv 13 Jul 2026.

## Placed in

- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 30 of 32)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [DocVQA: A Dataset for VQA on Document Images](https://mlsanity.com/paper/2007.00398/) (2020)
- [OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments](https://mlsanity.com/paper/2404.07972/) (2024)
- [τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environment](https://mlsanity.com/paper/2506.07982/) (2025)
- [API-Bank: A Comprehensive Benchmark for Tool-Augmented LLMs](https://mlsanity.com/paper/2304.08244/) (2023)
- [Executable Code Actions Elicit Better LLM Agents](https://mlsanity.com/paper/2402.01030/) (2024)
- [AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agents](https://mlsanity.com/paper/2405.14573/) (2024)
- [GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning](https://mlsanity.com/paper/2507.01006/) (2025)
- [WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models](https://mlsanity.com/paper/2401.13919/) (2024)
- [AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agents](https://mlsanity.com/paper/2407.18901/) (2024)
- [ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilities](https://mlsanity.com/paper/2408.04682/) (2024)
- [ScreenSpot-Pro: GUI Grounding for Professional High-Resolution Computer Use](https://mlsanity.com/paper/2504.07981/) (2025)
- [UI-TARS-2 Technical Report: Advancing GUI Agent with Multi-Turn Reinforcement Learning](https://mlsanity.com/paper/2509.02544/) (2025)
- [ChartQAPro: A More Diverse and Challenging Benchmark for Chart Question Answering](https://mlsanity.com/paper/2504.05506/) (2025)
- [MMSearch: Benchmarking the Potential of Large Models as Multi-modal Search Engines](https://mlsanity.com/paper/2409.12959/) (2024)
- [The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution](https://mlsanity.com/paper/2510.25726/) (2025)
- [GTA: A Benchmark for General Tool Agents](https://mlsanity.com/paper/2407.08713/) (2024)
- [MobileWorld: Benchmarking Autonomous Mobile Agents in Agent-User Interactive and MCP-Augmented Environments](https://mlsanity.com/paper/2512.19432/) (2025)
- [MCP-Atlas: A Large-Scale Benchmark for Tool-Use Competency with Real MCP Servers](https://mlsanity.com/paper/2602.00933/) (2026)
- [Ferret-UI 2: Mastering Universal User Interface Understanding Across Platforms](https://mlsanity.com/paper/2410.18967/) (2024)
- [MM-BrowseComp: A Comprehensive Benchmark for Multimodal Browsing Agents](https://mlsanity.com/paper/2508.13186/) (2025)
- [TIR-Bench: A Comprehensive Benchmark for Agentic Thinking-with-Images Reasoning](https://mlsanity.com/paper/2511.01833/) (2025)
- [Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments](https://mlsanity.com/paper/2602.11964/) (2026)
- [Beyond Seeing: Evaluating Multimodal LLMs on Tool-Enabled Image Perception, Transformation, and Reasoning](https://mlsanity.com/paper/2510.12712/) (2025)
- [OSWorld-MCP: Benchmarking MCP Tool Invocation In Computer-Use Agents](https://mlsanity.com/paper/2510.24563/) (2025)
- [ICDAR 2023 Competition on Hierarchical Text Detection and Recognition](https://mlsanity.com/paper/2305.09750/) (2023)
- [AgentVista: Evaluating Multimodal Agents in Ultra-Challenging Realistic Visual Scenarios](https://mlsanity.com/paper/2602.23166/) (2026)
- [Agent-X: Evaluating Deep Multimodal Reasoning in Vision-Centric Agentic Tasks](https://mlsanity.com/paper/2505.24876/) (2025)
- [Ferret-UI Lite: Lessons from Building Small On-Device GUI Agents](https://mlsanity.com/paper/2509.26539/) (2025)
- [ASTRA-bench: Evaluating Tool-Use Agent Reasoning and Action Planning with Personal User Context](https://mlsanity.com/paper/2603.01357/) (2026)

## Built on by (most cited first, 3 of 3)

- [Do Agent Benchmarks Do What They Say? An Executable-Contract Audit of Tool-Using Agent Environments](https://mlsanity.com/paper/2609.37315/) (2026)
- [MintAct: A Unified Visual Agent for Digital Environments](https://mlsanity.com/paper/2609.22083/) (2026)
- [ForkPilot: Self-Evolving Policy for Retrospective Search in Long-Horizon Agents](https://mlsanity.com/paper/2610.04889/) (2026)

## Abstract (arXiv)

We introduce MM-ToolSandBox, a benchmark and evaluation framework for visually grounded tool-calling agents. The framework provides a stateful execution environment spanning 500+ tools across 16 application domains, supporting multi-image, multi-turn tasks where agents must ground progressively arriving visual inputs into executable tool calls while handling realistic conversational phenomena (goal revisions, error corrections, state mutations). An automated scenario generation pipeline produces diverse, visually grounded scenarios through information-flow-guided planning and multi-stage quality filtering, yielding 258 human-verified nominal scenarios and 50 variants targeting interactive UI applications. Evaluating 12 state-of-the-art models, from 4B open-weight to frontier proprietary systems, shows that current models still lack robust visual tool-calling capability: even the best model achieves below 50% success rate. Our failure analysis further reveals that visual precision, not only planning, is a primary bottleneck for capable models: 53% of failures stem from incorrect information extraction from images despite otherwise correct task workflows. A planning-to-precision crossover emerges with scale: smaller models fail at deciding what to do, while larger models fail at perceiving what they see, suggesting fundamentally different research directions for improving models at different capability levels. The framework and the benchmark are publicly available at https://github.com/apple/ml-mmtoolsandbox