@inproceedings{wu-etal-2025-plot2code,
title = "{P}lot2{C}ode: A Comprehensive Benchmark for Evaluating Multi-modal Large Language Models in Code Generation from Scientific Plots",
author = "Wu, Chengyue and
Liang, Zhixuan and
Ge, Yixiao and
Guo, Qiushan and
Lu, Zeyu and
Wang, Jiahao and
Shan, Ying and
Luo, Ping",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-naacl.164/",
doi = "10.18653/v1/2025.findings-naacl.164",
pages = "3006--3028",
ISBN = "979-8-89176-195-7",
abstract = "Multi-modal Large Language Models have shown remarkable progress in visual contexts, yet their ability to convert visual figures into executable code remains underexplored. To address this, we introduce Plot2Code, a comprehensive benchmark designed to assess MLLMs' visual coding capabilities. Plot2Code includes 132 high-quality matplotlib plots across six plot types, as well as an additional 150 and 86 plots from Python{'}s and R{'}s plotly libraries respectively, totaling 368 plots. Each plot is paired with its source code and a descriptive instruction generated by GPT-4, enabling thorough evaluation across diverse inputs. Furthermore, we propose three automatic evaluation metrics{---}code pass rate, text-match ratio, and GPT-4V rating judgement{---}to assess the quality of generated code and rendered images. Notably, the GPT-4V rating demonstrates strong reliability, as it correlates well with human evaluations, particularly for datasets of a certain size. Cross-validation across MLLMs (GPT-4V, Gemini-1.5-Pro, and Claude-3-Opus) also shows high consistency in ratings, which likely stems from the fact that ratings are based on rendered images rather than direct MLLM outputs, indicating minimal bias for this metric. Our evaluation of 14 MLLMs, including both proprietary and open-source models, highlights significant challenges in visual coding, particularly for text-dense plots, where MLLMs heavily rely on textual instructions. We believe these findings will advance future development of MLLMs."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wu-etal-2025-plot2code">
<titleInfo>
<title>Plot2Code: A Comprehensive Benchmark for Evaluating Multi-modal Large Language Models in Code Generation from Scientific Plots</title>
</titleInfo>
<name type="personal">
<namePart type="given">Chengyue</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhixuan</namePart>
<namePart type="family">Liang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yixiao</namePart>
<namePart type="family">Ge</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qiushan</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zeyu</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiahao</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ying</namePart>
<namePart type="family">Shan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ping</namePart>
<namePart type="family">Luo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-195-7</identifier>
</relatedItem>
<abstract>Multi-modal Large Language Models have shown remarkable progress in visual contexts, yet their ability to convert visual figures into executable code remains underexplored. To address this, we introduce Plot2Code, a comprehensive benchmark designed to assess MLLMs’ visual coding capabilities. Plot2Code includes 132 high-quality matplotlib plots across six plot types, as well as an additional 150 and 86 plots from Python’s and R’s plotly libraries respectively, totaling 368 plots. Each plot is paired with its source code and a descriptive instruction generated by GPT-4, enabling thorough evaluation across diverse inputs. Furthermore, we propose three automatic evaluation metrics—code pass rate, text-match ratio, and GPT-4V rating judgement—to assess the quality of generated code and rendered images. Notably, the GPT-4V rating demonstrates strong reliability, as it correlates well with human evaluations, particularly for datasets of a certain size. Cross-validation across MLLMs (GPT-4V, Gemini-1.5-Pro, and Claude-3-Opus) also shows high consistency in ratings, which likely stems from the fact that ratings are based on rendered images rather than direct MLLM outputs, indicating minimal bias for this metric. Our evaluation of 14 MLLMs, including both proprietary and open-source models, highlights significant challenges in visual coding, particularly for text-dense plots, where MLLMs heavily rely on textual instructions. We believe these findings will advance future development of MLLMs.</abstract>
<identifier type="citekey">wu-etal-2025-plot2code</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.164</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.164/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>3006</start>
<end>3028</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Plot2Code: A Comprehensive Benchmark for Evaluating Multi-modal Large Language Models in Code Generation from Scientific Plots
%A Wu, Chengyue
%A Liang, Zhixuan
%A Ge, Yixiao
%A Guo, Qiushan
%A Lu, Zeyu
%A Wang, Jiahao
%A Shan, Ying
%A Luo, Ping
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Findings of the Association for Computational Linguistics: NAACL 2025
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-195-7
%F wu-etal-2025-plot2code
%X Multi-modal Large Language Models have shown remarkable progress in visual contexts, yet their ability to convert visual figures into executable code remains underexplored. To address this, we introduce Plot2Code, a comprehensive benchmark designed to assess MLLMs’ visual coding capabilities. Plot2Code includes 132 high-quality matplotlib plots across six plot types, as well as an additional 150 and 86 plots from Python’s and R’s plotly libraries respectively, totaling 368 plots. Each plot is paired with its source code and a descriptive instruction generated by GPT-4, enabling thorough evaluation across diverse inputs. Furthermore, we propose three automatic evaluation metrics—code pass rate, text-match ratio, and GPT-4V rating judgement—to assess the quality of generated code and rendered images. Notably, the GPT-4V rating demonstrates strong reliability, as it correlates well with human evaluations, particularly for datasets of a certain size. Cross-validation across MLLMs (GPT-4V, Gemini-1.5-Pro, and Claude-3-Opus) also shows high consistency in ratings, which likely stems from the fact that ratings are based on rendered images rather than direct MLLM outputs, indicating minimal bias for this metric. Our evaluation of 14 MLLMs, including both proprietary and open-source models, highlights significant challenges in visual coding, particularly for text-dense plots, where MLLMs heavily rely on textual instructions. We believe these findings will advance future development of MLLMs.
%R 10.18653/v1/2025.findings-naacl.164
%U https://aclanthology.org/2025.findings-naacl.164/
%U https://doi.org/10.18653/v1/2025.findings-naacl.164
%P 3006-3028
Markdown (Informal)
[Plot2Code: A Comprehensive Benchmark for Evaluating Multi-modal Large Language Models in Code Generation from Scientific Plots](https://aclanthology.org/2025.findings-naacl.164/) (Wu et al., Findings 2025)
ACL
- Chengyue Wu, Zhixuan Liang, Yixiao Ge, Qiushan Guo, Zeyu Lu, Jiahao Wang, Ying Shan, and Ping Luo. 2025. Plot2Code: A Comprehensive Benchmark for Evaluating Multi-modal Large Language Models in Code Generation from Scientific Plots. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 3006–3028, Albuquerque, New Mexico. Association for Computational Linguistics.