I am trying to implement some solution with my agent where the user send photos and a description and the agent have to recognize if the photo matches the description. I don’t know if are there any optimization in sending photos through the pass trough binary files.
Thanjs 
Hi @Gonzalo_Alegre
Do you have any more details about the workflow you are building?
This way we can better understand how to support you.
What is the initial input? How is the image shared? What model do you use? Etc.
Regards
I have the self hosted version of n8n. I am using the embedded chat with webhooks. I saved the images in azure and then in xano (firstly en azure because a need the link to store it in xano) and gpt 4o-mini
Great thanks!
Are the tools connected to the agent all functional and must they work in synergy? Because one suggestion might be to use an MPC to have a single tool control layer.
i don’t know if I could replace them because each is like a specialist in a field ( they are just a trigger and open ia node). The just give information for the principal agent
I hope you can help me
thanks
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.
