I hate that models are limited based on spending

I know that OpenAI limits the model availability based on usage spending but I’m not sure that I understand the reasoning for it.

It seems like you have to be between tier 3-5 to access an O3-mini assistant for example. In my case this is my personal OpenAI account and I work with AI most days at work at the moment (not on the same account). For me to POC or deliver projects I need access to the same models personally that I get professionally. This pattern also hinders very small one-man shops that want to leverage your API for their businesses.

I really like using OpenAI but I can’t see why I’d want to stick with using OpenAI over Genkit, Vertex, Gemini, etc if they provide superior models (to gpt-4o that I have access to in the API).

Anyone know if there’s plans to change this approach or is this just it?

Welcome to the community!

Models typically trickle down the tiers, so all you’d need to access o3 on tier 1 is probably patience.

Tier 3 is 100 bucks, but 100 bucks is a lot of money in low CoL countries. That said, tier 3 still had to wait several months before getting access to o1. o3 was faster, but likely because it’s a tiny model.

I don’t know if you’re missing out on all that much though. I don’t think that o3-mini will outperform o4-cot. O3 doesn’t come with any multimodal capabilities either, so there’s nothing you’re missing out on that front either.

Hope this helps somewhat :slight_smile:

1 Like