AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

TextMonkey: An OCR-Free Large Multimodal Model for Understanding Document

Yuliang Liu, Biao Yang, Qiang Liu, Zhang Li, Zhiyin Ma, Shuo Zhang, Xiang Bai

TPAMI 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Interleaved image-text generation with MLLMs · 243 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add zero initialization to shifted window attention (ablation)TextMonkey · ChartQAaccuracy: 57.6 -> 58.2Improved
add shifted window attention (cross-window connections) (ablation)TextMonkey · DocVQA / TextVQAaccuracy: 62.6 -> 64.1 DocVQA; 62.4 -> 64.8 TextVQA; 45.9 -> 46.0 SROIEImproved
add token resampler on top of window attention (ablation)TextMonkey · SROIE / DocVQA / TextVQAaccuracy: 46.0 -> 47.0 SROIE; 64.1 -> 64.3 DocVQA; 64.8 -> 65.9 TextVQAImproved
use significant-token filter instead of random queries (ablation)Token Resampler · SROIE / DocVQA / TextVQAaccuracy: 32.9 -> 47.0 SROIE; 46.7 -> 64.3 DocVQA; 59.5 -> 65.9 TextVQAImproved
add resampler after token filtering instead of dropping tokens (ablation)Token Resampler · SROIE / DocVQA / TextVQAaccuracy: 44.9 -> 47.0 SROIE; 63.5 -> 64.3 DocVQA; 62.5 -> 65.9 TextVQAImproved
sort selected tokens by order (ablation)Token Resampler · DocVQAaccuracy: 62.1 -> 64.3Improved
raise resolution from 896 to 1344 without token compression (ablation)TextMonkey · DocVQAaccuracy: 64.1 -> 54.9Hurt
raise resolution to 1344 with 1024 kept tokens vs 896 with 512 (ablation)TextMonkey · InfoVQA / DocVQAaccuracy: 28.2 -> 31.4 InfoVQA; 64.3 -> 64.5 DocVQA; 65.9 -> 65.1 TextVQA; 47.0 -> 46.0 SROIEMixed
require model to output answer position (ablation)TextMonkey · DocVQA / SROIEaccuracy: 64.3 -> 64.5 DocVQA; 47.0 -> 47.2 SROIEImproved
require model to output answer position (ablation)TextMonkey · ChartQA / InfoVQAaccuracy: 58.2 -> 57.8 ChartQA; 28.2 -> 27.7 InfoVQAHurt
use point instead of rectangle position representation (ablation)TextMonkey · DocVQA / SROIEaccuracy: 64.3 -> 65.0 DocVQA; 47.0 -> 47.9 SROIEImproved
use polygon instead of rectangle position representation (ablation)TextMonkey · DocVQA / SROIEaccuracy: 64.3 -> 64.0 DocVQA; 47.0 -> 47.2 SROIEHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy