PulseAugur
EN
LIVE 17:35:22

New benchmark reveals MLLMs struggle with visual-to-code reproduction

By PulseAugur Editorial · [1 sources] ·

A new benchmark called FigCodeBench has been developed to evaluate the capabilities of Multimodal Large Language Models (MLLMs) in reproducing complex visual figures and generating corresponding code. This framework addresses the gap in current benchmarks by integrating visual understanding and code generation, moving beyond isolated assessments. Experiments conducted on 24 proprietary and open-source MLLMs, including Gemini-3.1 Pro and GPT-5.4, revealed a significant performance drop across various programming languages and difficulty levels, offering insights into their limitations. AI

IMPACT Highlights limitations in current MLLMs for complex visual-to-code tasks, potentially guiding future model development.

RANK_REASON New academic paper introducing a novel benchmark and evaluation framework for MLLMs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

New benchmark reveals MLLMs struggle with visual-to-code reproduction

How we ranked this

Signal score
4 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
New academic paper introducing a novel benchmark and evaluation framework for MLLMs. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Same-day
Cluster formed today. Ranking reflects the current source set at time of score.

Full methodology in our editorial standards.

COVERAGE [1]

  1. arXiv cs.CV TIER_1 English(EN) · Zijian Chen, Zhengyu Chen, Bohan Liang, Lirong Deng, Yushuo Zheng, Yanwei Jiang, Qi Jia, Kaiwei Zhang, Wenjun Zhang, Guangtao Zhai ·

    From Pixel to Coding: Evaluating the Figure Reproduction Capabilities of MLLMs

    arXiv:2610.10066v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities in both visual understanding and code generation. However, existing benchmarks typically evaluate these two modalities in isolation, lacking a dedica…

RELATED ENTITIES

RELATED TOPICS