Olmocr 7B Thai V2
O
Olmocr 7B Thai V2
Developed by Adun
The optimized olmOCR model focuses on improving the accuracy of Thai text recognition and supports multiple languages and table formats.
Downloads 917
Release Time : 4/21/2025
Model Overview
olmOCR is an OCR tool based on the vision-language model, which has been fine-tuned to enhance the recognition ability of Thai characters and numbers and is suitable for extracting text from documents such as PDFs.
Model Features
Multilingual and Table Support
Supports the recognition of characters in multiple languages and table formats.
Open-source Feature
Provides model weights, fine-tuning datasets, and inference code to facilitate developers' customized development.
High Accuracy
Fine-tuned based on 250K documents to ensure recognition accuracy.
API and CLI Support
Can be called via the command line or API (vLLM, SGlang) for easy integration into existing systems.
Model Capabilities
Thai Text Recognition
Multilingual Character Recognition
Table Format Recognition
PDF Text Extraction
Use Cases
Document Processing
Digitization of Thai Documents
Convert Thai PDF documents into editable plain text.
Improve the accuracy of Thai character recognition.
Multilingual Table Recognition
Extract structured data from documents containing multiple languages and tables.
Support complex document formats.
Featured Recommended AI Models
Š 2025AIbase