[GH-ISSUE #868] Enhance Image Understanding: Let DeepSeek Read All Types of Images! #512

New issue

Closed

opened 2026-10-26 21:22:32 +00:00 by bsakaguchi · 3 comments

bsakaguchi commented

2026-10-26 21:22:32 +00:00

Owner

Originally created by @Nero007x on GitHub (May 11, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-V3/issues/868

Is your feature request related to a problem? Please describe.
Yeah, kind of. I get frustrated when I upload images that aren’t just text—like graphs, photos, or diagrams—and DeepSeek doesn’t really do much with them. It feels like it only "gets" images that have clear text, and ignores everything else.

Describe the solution you'd like
It’d be awesome if DeepSeek could understand all kinds of images, not just the ones with text. Like, if I upload a chart, it could explain what’s going on. Or if it’s a photo, it could describe the scene, objects, or even emotions. Just more flexibility with visuals in general.

Describe alternatives you've considered
I’ve tried other tools like GPT-4 Vision, which does a better job with this. But I’d really prefer to stick with DeepSeek if it had the same kind of image understanding.

Additional context
This could open up a bunch of new use cases—like design feedback, analyzing screenshots or app mockups, even reading messy whiteboard photos. Maybe there could even be a setting to choose between “text-only” and “full image understanding” modes.

Originally created by @Nero007x on GitHub (May 11, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-V3/issues/868 **Is your feature request related to a problem? Please describe.** Yeah, kind of. I get frustrated when I upload images that aren’t just text—like graphs, photos, or diagrams—and DeepSeek doesn’t really do much with them. It feels like it only "gets" images that have clear text, and ignores everything else. **Describe the solution you'd like** It’d be awesome if DeepSeek could understand all kinds of images, not just the ones with text. Like, if I upload a chart, it could explain what’s going on. Or if it’s a photo, it could describe the scene, objects, or even emotions. Just more flexibility with visuals in general. **Describe alternatives you've considered** I’ve tried other tools like GPT-4 Vision, which does a better job with this. But I’d really prefer to stick with DeepSeek if it had the same kind of image understanding. **Additional context** This could open up a bunch of new use cases—like design feedback, analyzing screenshots or app mockups, even reading messy whiteboard photos. Maybe there could even be a setting to choose between “text-only” and “full image understanding” modes. <img width="842" alt="Image" src="https://github.com/user-attachments/assets/e0986078-b2c4-49f8-861c-cb00de5078e1" />

bsakaguchi

2026-10-26 21:22:32 +00:00

closed this issue
added the
stale

closed-as-stale
labels

bsakaguchi commented

2026-10-26 21:22:33 +00:00

Author

Owner

@github-actions[bot] commented on GitHub (Jun 11, 2025):

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!

@github-actions[bot] commented on GitHub (Jun 11, 2025): This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!

bsakaguchi commented

2026-10-26 21:22:33 +00:00

Author

Owner

@github-actions[bot] commented on GitHub (Jun 26, 2025):

false

@github-actions[bot] commented on GitHub (Jun 26, 2025): false

bsakaguchi commented

2026-10-26 21:22:33 +00:00

Author

Owner

@adoudathereal commented on GitHub (Jul 26, 2025):