Is it yet possible to include images in the input when playing around with the API? I know it’s possible with the public facing chatGPT.
At present, the GPT-4 Vision (GPT4V) API has not been made available to the public, meaning that you cannot use GPT-4 for image input-related tasks. However, there is a possibility that the GPT-4 Vision API will be launched for public use during OpenAI Dev Day. You can wait for it or you can build with other API’s like Bing and Google.
Thanks I though that would be the case. Will watch out for Dev Day
That’s great in the meantime you can use open-source models such as LLaVA, which is also effective.
How to make an application with LLaVA? Is there an API?
You can use Llamaindex they just release multi-modal , they are currently using replicate to access LLaVA.
Thank you for the explanation!
I will have a look.