# Texo: Formula Recognition within 20M Parameters

arXiv: https://arxiv.org/abs/2602.17189 · HTML: https://mlsanity.com/paper/2602.17189/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 37 references that are in the map.
- First posted on arXiv 19 Feb 2026.

## Builds on (most cited first, 27 of 27)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models](https://mlsanity.com/paper/2304.10592/) (2023)
- [How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites](https://mlsanity.com/paper/2404.16821/) (2024)
- [Multilingual Denoising Pre-training for Neural Machine Translation](https://mlsanity.com/paper/2001.08210/) (2020)
- [DETRs Beat YOLOs on Real-time Object Detection](https://mlsanity.com/paper/2304.08069/) (2023)
- [OCR-free Document Understanding Transformer](https://mlsanity.com/paper/2111.15664/) (2021)
- [Nougat: Neural Optical Understanding for Academic Documents](https://mlsanity.com/paper/2308.13418/) (2023)
- [DeepSeek-OCR: Contexts Optical Compression](https://mlsanity.com/paper/2510.18234/) (2025)
- [General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model](https://mlsanity.com/paper/2409.01704/) (2024)
- [OmniDocBench: Benchmarking Diverse PDF Document Parsing with Comprehensive Annotations](https://mlsanity.com/paper/2412.07626/) (2024)
- [RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer](https://mlsanity.com/paper/2407.17140/) (2024)
- [Image-to-Markup Generation with Coarse-to-Fine Attention](https://mlsanity.com/paper/1609.04938/) (2016)
- [Understanding LLMs: A Comprehensive Overview from Training to Inference](https://mlsanity.com/paper/2401.02038/) (2024)
- [PaddleOCR-VL-1.5: Towards a Multi-Task 0.9B VLM for Robust In-the-Wild Document Parsing](https://mlsanity.com/paper/2601.21957/) (2026)
- [Fast Vocabulary Transfer for Language Model Compression](https://mlsanity.com/paper/2402.09977/) (2024)
- [UniMERNet: A Universal Network for Real-World Mathematical Expression Recognition](https://mlsanity.com/paper/2404.15254/) (2024)
- [Image Over Text: Transforming Formula Recognition Evaluation with Character Detection Matching](https://mlsanity.com/paper/2409.03643/) (2024)
- [When Counting Meets HMER: Counting-Aware Network for Handwritten Mathematical Expression Recognition](https://mlsanity.com/paper/2207.11463/) (2022)
- [Handwritten Mathematical Expression Recognition with Bidirectionally Trained Transformer](https://mlsanity.com/paper/2105.02412/) (2021)
- [Improving Attention-Based Handwritten Mathematical Expression Recognition with Scale Augmentation and Drop Attention](https://mlsanity.com/paper/2007.10092/) (2020)
- [Handwritten Mathematical Expression Recognition via Attention Aggregation based Bi-directional Mutual Learning](https://mlsanity.com/paper/2112.03603/) (2021)
- [HunyuanOCR Technical Report](https://mlsanity.com/paper/2511.19575/) (2025)
- [Fine-Tuning Transformers: Vocabulary Transfer](https://mlsanity.com/paper/2112.14569/) (2021)
- [Pattern Generation Strategies for Improving Recognition of Handwritten Mathematical Expressions](https://mlsanity.com/paper/1901.06763/) (2019)
- [PP-FormulaNet: Bridging Accuracy and Efficiency in Advanced Formula Recognition](https://mlsanity.com/paper/2503.18382/) (2025)

## Abstract (arXiv)

In this paper we present Texo, a minimalist yet highperformance formula recognition model that contains only 20 million parameters. By attentive design, distillation and transfer of the vocabulary and the tokenizer, Texo achieves comparable performance to state-of-the-art models such as UniMERNet-T and PPFormulaNet-S, while reducing the model size by 80% and 65%, respectively. This enables real-time inference on consumer-grade hardware and even in-browser deployment. We also developed a web application to demonstrate the model capabilities and facilitate its usage for end users.