Open hongkongkiwi opened 1 month ago
hey @hongkongkiwi , we don't currently support images at this point (we started this project just a few weeks ago), but definitely something we wanna do in the future.
Hello! We do support image recognition for supported models like gpt-4o
, but the source image must be currently hosted somewhere else. Adding an image
tag inside a user
message will let the LLM analyse the contents of the image:
---
provider: Latitude
model: gpt-4o
---
Describe what you can see in this image:
<user>
<image>{{image_url}}</image>
</user>
Does model support vision/images ?
This is a key feature for me and I'd love to be able to test some prompts with images.