Install MiniCPM Llama3-V 2.5 Locally - Beats GPT4o 

You are so welcome! Much appreciated.

agreed

That's a very good request

Maybe it's possible to do that locally, there should be a script or tool for ollama to do this if I remember correctly. Once it can be loaded in Ollama it can be used with Open WebUI one can just drag/drop images there. We'll see. :)

Noted.

in the Model card they wrote (for OCR Capabilities): 💪 Strong OCR Capabilities. MiniCPM-Llama3-V 2.5 can process images with any aspect ratio up to 1.8 million pixels, achieving an 700+ score on OCRBench, surpassing proprietary models such as GPT-4o, GPT-4V-0409, Qwen-VL-Max and Gemini Pro. Based on recent user feedback, MiniCPM-Llama3-V 2.5 has now enhanced full-text OCR extraction, table-to-markdown conversion, and other high-utility capabilities, and has further strengthened its instruction-following and complex reasoning abilities, enhancing multimodal interaction experiences. I could not find the OCRBench score for GPT-4o !! may be it was tested by the model developers!! but the evaluation comaprison on the photo showing that it is a good model: (It was compaird with GPT4-v-1106): huggingface.co/openbmb/MiniCPM-Llama3-V-2_5/resolve/main/assets/MiniCPM-Llama3-V-2.5-peformance.png

On some benchmarks, plz see the model card for the numbers.

would need to check

use lm studio

yes

Noted

Both are VLMs

May be from huggingface