Agent-X: Evaluating Deep Multimodal Reasoning in Vision-Centric Agentic Tasks

ICLR 2026 Conference Submission15196 Authors

Keywords: Vision centric Agents, Deep Reasoning, VLMs, Tool Use Evaluation

TL;DR: Agent-X benchmark (828 multimodal tasks) tests deep, stepwise reasoning in vision agents; SOTA models score <50%, revealing major gaps in multimodal reasoning and tool use.

Abstract: Deep reasoning is fundamental for solving complex tasks, especially in vision-centric scenarios that demand sequential, multimodal understanding. However, existing benchmarks typically evaluate agents with fully synthetic, single-turn queries, limited visual modalities, and lack a framework to assess reasoning quality over multiple steps as required in real-world settings. To address this, we introduce Agent-X, a large-scale benchmark for evaluating vision-centric agents’ multistep and deep reasoning capabilities in real-world, multimodal settings. AgentX features 828 agentic tasks with authentic visual contexts, including images, multi-image comparisons, videos, and instructional text. These tasks span six major agentic environments: general visual reasoning, web browsing, security and surveillance, autonomous driving, sports, and math reasoning. Our benchmark requires agents to integrate tool use with explicit, stepwise decision-making in these diverse settings. In addition, we propose a fine-grained, step-level evaluation framework that assesses the correctness and logical coherence of each reasoning step and the effectiveness of tool usage throughout the task. Our results reveal that even the best-performing models, including GPT, Gemini, and Qwen families, struggle to solve multi-step vision tasks, achieving less than 50% full-chain success. These findings highlight key bottlenecks in current LMM reasoning and tool-use capabilities and identify future research directions in vision-centric agentic reasoning models

Primary Area: datasets and benchmarks

Submission Number: 15196