Hello,
I’m trying to look into tools that are using GPT4, however I can’t seem to generate anything but a GPT3.5 key, which, oddly enough, is rate-limited at its first use, even though I’m a GPT4plus subscriber for a couple months now AND have added a 5$ pre-paid balance. What gives?
All my keys start with ‘sk-’ and I have no options to change the model anywhere.
Hi and welcome to the Developer Forum!
If you have added $5 of prepay credit to your API account you should now be able to generate a new API key which has GPT-4 access enabled.
It doesn’t happen, apps report stuff such as:
“openai.error.InvalidRequestError: The model gpt-4 does not exist or you do not have access to it. Learn more: How can I access GPT-4? | OpenAI Help Center”
The keys I can generate now look identical to the ones generated previously - I have no dropdown and no way to select any type of key and I’m unsure whether that’s normal or not.
The payment to your account may take a little time to register throughout systems.
The primary place to check is the API Playground. See if under chat mode, you have gpt-4 as a model that can be selected.
Make sure under the API key section, you have selected the correct default organization at the bottom of that screen.
I do have CHAT4 options at OpenAI Platform
On OpenAI Platform however, on my lone ‘organization’, there’s no options still for the type of key I get to create.
I’ll wait some more, but I feel like the option won’t appear anytime soon.
The only place in your account that will show gpt-4 has been enabled is within rate limits. It might look like this:
Rate limits
OpenAI enforces rate limits on the requests you can make to the API. These are applied over tokens-per-minute, requests-per-minute (in some cases requests-per-day), or in the case of image models, images-per-minute.
Learn more in the rate limits documentation, or reference the default rate limits for our models. Your organization’s rate limits are listed below.
Model TPM RPM Chat gpt-3.5-turbo 90,000 3,500 gpt-4 10,000 200 gpt-4-0314 10,000 200 gpt-4-0613 10,000 200
Looks like it’s working now, thanks guys! Albeit not much difference than what I was seeing before, the ‘rate limits’ page shows:
|MODEL|TPM|RPM|
| --- | --- | --- |
|CHAT|
|gpt-3.5-turbo|90,000|3,500|
|gpt-3.5-turbo-0301|90,000|3,500|
|gpt-3.5-turbo-0613|90,000|3,500|
|gpt-3.5-turbo-16k|180,000|3,500|
|gpt-3.5-turbo-16k-0613|180,000|3,500|
|gpt-4|10,000|200|
|gpt-4-0314|10,000|200|
|gpt-4-0613|10,000|200|
I just wanted to mention that this thread got me to delete my existing payment plan (pay as you go). When I added a new one, it was automatically set to the prepay plan, and when that had $5 in it, I finally had gpt4 access even with my existing key.
Whew. I’m glad I don’t have to worry about that anymore. Now I just have to watch my spending. 
So if I have a payment plan of $20.00 a month, this still does not mean I automatically get access to API keys. I actually need to be actively generating say 200 plus calls say on a monthly basis or is that a daily or weekly basis to get the API key?
You are correct, that ChatGPT’s $20 subscription doesn’t affect an API account.
However, you are mistaken about the requirement for some other qualification. API is a pay per use service, and its access portal is at platform.openai.com. An account may have a trial credit (or it may have expired), and then usage credits can be purchased to load the account. API keys can be generated any time - you just might not have a funding method available when you try to use them.
And, if you don’t know what its for…you shouldn’t use it. There are many scam apps and extensions that would like you to put in an API key (which should be used just by developers on their own server), and then your key can be stolen and account emptied by nefarious use.
Yeah - but the support docs describe “GPT4 - instant access - here’s how”, and not “generate 200 calls and check back in a month”
API keys are general, meaning one API key can be used for GPT 3.5 TURBO, GPT 4, Dall-e etc.
If you don’t wanna use GPT 3.5 you can specify this in your code while developing application.
{
"model": "gpt-4-1106-preview",
"messages": [
{"role": "user", "content": "<prompt-start> <prompt-before-input1> <input1> <prompt-before-input2><input2> <prompt-before-input3> <input3> <prompt-before-input4> <input4> <prompt-before-input5> <input5>"}
],
"max_tokens": <max-tokens>,
"n": <number-of-renditions>
}
Under “model” you can specify gpt-4-1106-preview or any other model you are using.
In your usage tab should appear GPT 4 usage. Hope this helps. 