I am using the chat completions API with gpt-4-vision-preview. It works great for my jpeg and png images, but I notice the API does not support .bmp images.
Is there any straightforward way I can use .bmp images? I know I can manually convert and re-upload or write a python script. But I would prefer not to do that given I already have the image URLs.
I ended up using PIL to convert to png, and then uploading to an S3 bucket. A little annoying but it worked
There’s a quick answer in the vision guide:
### What type of files can I upload?
We currently support PNG (.png), JPEG (.jpeg and .jpg), WEBP (.webp), and non-animated GIF (.gif).
### Is there a limit to the size of the image I can upload?
Yes, we restrict image uploads to 20MB per image.
Preprocessing the images for vision not only lets you ensure a wide variety can be sent, but you can also control costs, understanding the “tiles” pricing structure, and for example keeping the largest dimension under 1024px for maximum 4 tiles, or 512px for just one.
Images can be sent directly in the request by conversion to UTF-8 base64.