Pricing introduction killed dalle2 prompt exploration for me, even though I could afford playing with it. The feeling of wasted time and money for getting garbage results makes it just not worth it. I haven't been excited to explore a prompt idea since

> Hello, I am Eliza. 'if you’re an openai api customer, i’d love your feedback.' > Why are you concerned about my feedback? 'if you looked at using the api and didn’t, why not?' > We were discussing you, not me.

What annoys me the most is 1. Prompt writing - the data preparation before sending it to the completions endpoint because one char can change output - missing auto removing new lines. 2. Missing structured output.

As a Indian Its very expensive for me, I used the free version, and it's amazing. I wish it could be free. 😅

Make GPT4

1. The rate limiting. It prevents us to offer our app to more customers. We were planning to contact openai about that. 2. The 4k tokens limit on GPT3. We use it to offer data-extaction like functionality, but it puts a hard limit on document sizes 👇

When is GPT4 launching though

I don't think the API likes me:

I'd like an explicit command to see if a model has been loaded and is ready for inference. Right now I send a request and have custom logic based on the response

still waiting on access to Dalle-2