Enable Image Generation

agentcoinorg / evo.ninja

A versatile generalist agent.

MIT License

1.05k stars 174 forks source link

Enable Image Generation #465

Open rihp opened 8 months ago

rihp commented 8 months ago

https://platform.openai.com/docs/guides/images/example-dall-e-3-generations?context=node

dOrgJelli commented 7 months ago

My questions & thoughts on prioritizing this:

If we have significant demand for this, we should explore.
We need to understand what the cost of this has, and if it works well with our user subsidies as they function right now.
What are the use-cases for this? What users have requested this?

rihp commented 7 months ago

So far 1 user (kev on discord) has requested image and video generation

We can integrate Dall-e 3 which costs about this much Development time shouldnt take more than a couple of days afaik

The usecase is writing marketing articles and tiktok scripts, the images are then processed into video manually (or programatically) and the user narrates the script to create the tiktok video

orishim commented 7 months ago

I would not prioritize image generation for now, since we are already working on significant new functionality that will be quite complex and novel and does not relate to media generation capabilities. I think this would distract us at the current moment, but we should revisit in a few weeks depending on how that goes, and if further requests come in for this feature.