Will GeminiAI read websites which has images embedded in it. The images contains text or code. if yes how to ask geminiai to read the website with image and provide summary of it.
Welcome to the forums!
The Gemini API and AI Studio won’t read a website.
You, as a developer, can build something that does read a website and feeds it to the Gemini API, but that’s different. You’ll control how to handle images and any other text it encounters.
Yes it can atleast my project can it
Hi @David_Wieczorek,
Prompts play a big role in how LLMs respond. In the below instance, Gemini’s answer was direct, but the way we phrase our questions can significantly influence the results we get from large language models.
1 Like
Well i understand that tho gemini is indead capable of it as well as having agents and multiple servers depends how much you know about LLMs and how you work with them
Remember that Large Language Models are not sources of truth. Just because they say they can do something - doesn’t mean they can.
They are pattern machines that are designed to present reasonable continuations of what they are prompted, so the prompt that you setup encouraged it to reply that it could. So it said it could.
Did you actually tell it to read from the website?
Here’s what I got when I did.
Before yes , also same us people say you cant not have multiple agents tho here i am with multiple agents system thats work just fine for research