How do you make your banner images? I'm assuming some AI tool, but how do you get it to remember your character? (also... good luck in the post mines XD...)
My favorite is still the *old* OpenAI Sora interface, using GPT-Image-1.5, (only accessible with a European IP since the new Sora is not GDPR compliant or something, but that interface doesn't have the option for image gen).
You can add up to 10 images as context and the model infers details from that.
It's pretty good on the pareto of [style] + [being creative] + [instruction following] + [character consistency]. It's not perfect at any of these but it's the only one that is sufficiently good at all these for me.
Sadly, the interface will be deprecated in 2 weeks and I'll need to switch to a different UI. You can use the same model through normal ChatGPT chats, but it only generates one image at a time and is quite slow and buggy and hard to sort through.
How do you make your banner images? I'm assuming some AI tool, but how do you get it to remember your character? (also... good luck in the post mines XD...)
My favorite is still the *old* OpenAI Sora interface, using GPT-Image-1.5, (only accessible with a European IP since the new Sora is not GDPR compliant or something, but that interface doesn't have the option for image gen).
You can add up to 10 images as context and the model infers details from that.
It's pretty good on the pareto of [style] + [being creative] + [instruction following] + [character consistency]. It's not perfect at any of these but it's the only one that is sufficiently good at all these for me.
Sadly, the interface will be deprecated in 2 weeks and I'll need to switch to a different UI. You can use the same model through normal ChatGPT chats, but it only generates one image at a time and is quite slow and buggy and hard to sort through.