[GH-ISSUE #12145] Support for BGE-VL-Screenshot Model as Image Vector Embedding Model #54587

Closed
opened 2026-04-29 06:27:18 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @AuditAIH on GitHub (Sep 2, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12145

Support for BGE-VL-Screenshot Model as Image Vector Embedding Model
from:https://huggingface.co/mradermacher/BGE-VL-Screenshot-GGUF

Dear Team,

I would like to propose adding support for the BGE-VL-Screenshot model as an option for image vector embedding in our system.

This model, available at [URL], shows promising capabilities in generating high-quality vector embeddings from images. Integrating it would expand our range of image embedding options and potentially provide better performance for certain use cases, particularly those involving screenshot images.

Key reasons for supporting this model:

  1. It specializes in handling screenshot images, which is a common use case in our application
  2. It can generate meaningful vector representations that capture visual content effectively
  3. Adding it would give users more choices for image embedding tasks

Could we please evaluate the feasibility of integrating the BGE-VL-Screenshot model into our framework? I'm happy to provide further information or assist with testing if needed.

Thank you for considering this request.

Best regards

Image
Originally created by @AuditAIH on GitHub (Sep 2, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12145 Support for BGE-VL-Screenshot Model as Image Vector Embedding Model from:https://huggingface.co/mradermacher/BGE-VL-Screenshot-GGUF Dear Team, I would like to propose adding support for the BGE-VL-Screenshot model as an option for image vector embedding in our system. This model, available at [[URL]](https://modelscope.cn/models/dengcao/BGE-VL-Screenshot-GGUF/files), shows promising capabilities in generating high-quality vector embeddings from images. Integrating it would expand our range of image embedding options and potentially provide better performance for certain use cases, particularly those involving screenshot images. Key reasons for supporting this model: 1. It specializes in handling screenshot images, which is a common use case in our application 2. It can generate meaningful vector representations that capture visual content effectively 3. Adding it would give users more choices for image embedding tasks Could we please evaluate the feasibility of integrating the BGE-VL-Screenshot model into our framework? I'm happy to provide further information or assist with testing if needed. Thank you for considering this request. Best regards <img width="2245" height="525" alt="Image" src="https://github.com/user-attachments/assets/692d55f8-58b3-4249-ae86-09facb7e1862" />
GiteaMirror added the model label 2026-04-29 06:27:18 -05:00
Author
Owner

@rick-github commented on GitHub (Sep 2, 2025):

#5304

<!-- gh-comment-id:3243511718 --> @rick-github commented on GitHub (Sep 2, 2025): #5304
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#54587