yongxianwei/Qwen2.5-VL-32B-OCR
VISIONPricing:Input $1.06 / Cached $0.053 / Output $2.6Concurrent Unit Cost:2Model Size:32BQuant:FP8Context Size:32kPublished:Nov 18, 2025License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold
yongxianwei/Qwen2.5-VL-32B-OCR is a 32 billion parameter vision-language model, based on Qwen/Qwen2.5-VL-32B, specifically fine-tuned for Optical Character Recognition (OCR) tasks. This LoRA adapter enhances the base model's ability to accurately extract text from images. It is designed for applications requiring robust text recognition from visual inputs.
Loading preview...
Overview
This model, yongxianwei/Qwen2.5-VL-32B-OCR, is a specialized LoRA (Low-Rank Adaptation) adapter built upon the powerful Qwen/Qwen2.5-VL-32B base model. It has been meticulously fine-tuned to excel in Optical Character Recognition (OCR) tasks, making it highly effective for extracting text from various image formats.
Key Capabilities
- Enhanced OCR Performance: Significantly improves the base model's ability to recognize and extract text from images.
- Vision-Language Integration: Leverages the robust vision-language understanding of the Qwen2.5-VL-32B architecture.
- Efficient Adaptation: Utilizes LoRA for efficient fine-tuning, allowing for specialized performance without retraining the entire 32 billion parameter base model.
Good For
- Text Extraction: Ideal for applications requiring accurate text recognition from scanned documents, images, or other visual sources.
- Document Processing: Can be integrated into workflows for automating data entry, digitizing physical records, or analyzing image-based content.
- Research and Development: Provides a strong foundation for further experimentation and fine-tuning on specific OCR datasets or challenging visual text scenarios.