@inproceedings{li-etal-2025-text,
title = "Text or Pixels? Evaluating Efficiency and Understanding of {LLM}s with Visual Text Inputs",
author = "Li, Yanhong and
Lan, Zixuan and
Zhou, Jiawei",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.558/",
doi = "10.18653/v1/2025.findings-emnlp.558",
pages = "10564--10578",
ISBN = "979-8-89176-335-7",
abstract = "Large language models (LLMs) and their multimodal variants can now process visual inputs, including images of text. This raises an intriguing question: Can we compress textual inputs by feeding them as images to reduce token usage while preserving performance?In this paper, we show that \textit{visual text representations} are a practical and surprisingly effective form of input compression for decoder LLMs. We exploit this idea by rendering long text inputs as a single image and providing it directly to the model. This approach dramatically reduces the number of decoder tokens required, offering a new form of input compression. Through experiments on two distinct benchmarks {---} RULER (long-context retrieval) and CNN/DailyMail (document summarization) {---} we demonstrate that this text-as-image method yields substantial token savings \textit{without degrading task performance}."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2025-text">
<titleInfo>
<title>Text or Pixels? Evaluating Efficiency and Understanding of LLMs with Visual Text Inputs</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yanhong</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zixuan</namePart>
<namePart type="family">Lan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiawei</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-335-7</identifier>
</relatedItem>
<abstract>Large language models (LLMs) and their multimodal variants can now process visual inputs, including images of text. This raises an intriguing question: Can we compress textual inputs by feeding them as images to reduce token usage while preserving performance?In this paper, we show that visual text representations are a practical and surprisingly effective form of input compression for decoder LLMs. We exploit this idea by rendering long text inputs as a single image and providing it directly to the model. This approach dramatically reduces the number of decoder tokens required, offering a new form of input compression. Through experiments on two distinct benchmarks — RULER (long-context retrieval) and CNN/DailyMail (document summarization) — we demonstrate that this text-as-image method yields substantial token savings without degrading task performance.</abstract>
<identifier type="citekey">li-etal-2025-text</identifier>
<identifier type="doi">10.18653/v1/2025.findings-emnlp.558</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.558/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>10564</start>
<end>10578</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Text or Pixels? Evaluating Efficiency and Understanding of LLMs with Visual Text Inputs
%A Li, Yanhong
%A Lan, Zixuan
%A Zhou, Jiawei
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F li-etal-2025-text
%X Large language models (LLMs) and their multimodal variants can now process visual inputs, including images of text. This raises an intriguing question: Can we compress textual inputs by feeding them as images to reduce token usage while preserving performance?In this paper, we show that visual text representations are a practical and surprisingly effective form of input compression for decoder LLMs. We exploit this idea by rendering long text inputs as a single image and providing it directly to the model. This approach dramatically reduces the number of decoder tokens required, offering a new form of input compression. Through experiments on two distinct benchmarks — RULER (long-context retrieval) and CNN/DailyMail (document summarization) — we demonstrate that this text-as-image method yields substantial token savings without degrading task performance.
%R 10.18653/v1/2025.findings-emnlp.558
%U https://aclanthology.org/2025.findings-emnlp.558/
%U https://doi.org/10.18653/v1/2025.findings-emnlp.558
%P 10564-10578
Markdown (Informal)
[Text or Pixels? Evaluating Efficiency and Understanding of LLMs with Visual Text Inputs](https://aclanthology.org/2025.findings-emnlp.558/) (Li et al., Findings 2025)
ACL