I use OpenAI for writing and researching. I don’t use large datasets, databases, etc. I see an API is good for privacy (but I don’t need that). I also see API gives me real-time data, which would be useful for my work.
- Is there any advantage for me to use the API to connect to OpenAI?
- How does the API access affect the way I use OpenAI ChatGPT.com? Would I use a different access?
- Is API usage something only for developers?
- Is anyone using API access for personal use of ChatGPT.com? Why? How does it change the way you work? Advantages? Disadvantages?
2 Likes
Hi @AndreasRamos and welcome back!
Let me try to answer your questions:
For building apps/services it is the only way (either via OpenAI API directly, or via Microsoft Azure API for example - if you have very strict enterprise requirements, such as uptime, SLAs, etc).
It is also great for building your own personal ChatGPT (e.g. by combining the API with something like Open WebUI) - because it lets you have a fairly raw experience directly with LLMs - ChatGPT implements lots of things behind the scenes taking away significant control from your hands.
For API you basically need to buy credits. At the minimum you need to pay 5 USD, and you are charged on how many input and output tokens are consumed. So it’s decoupled from ChatGPT , which is a subscription based service.
Not necessarily - depends on how savvy or interested you are in building some minimalistic services - I know plenty of investment people, lawyers and property managers who have build amazing tools using the API.
I know people that have coupled it with Open WebUI (disclaimer: I haven’t!) and they usually have several API providers and rotate between them. So that’s the main advantage - that you can have a single interface for different model providers. At the same time, you lose lots of additional functionality, like realtime voice, deep research, canvas, etc.
4 Likes
That’s precisely what I use the most, deep research. Many thanks for your detailed reply. I will pass this on.
2 Likes