Incomprehensible cost of tokens when accessing the API

[versed rapids]

For some reason, when accessing the /v1/chat/completions endpoint, I encounter a very high cost of requests.
Request:
{
"model": "gpt-3.5-turbo",
"messages": [{"role": "user", "content": "Hi what's up?"}]
}
Response:
"created": 1681723705,
"model": "gpt-3.5-turbo-0301",
"usage": {
"prompt_tokens": 13,
"completion_tokens": 36,
"total_tokens": 49
},
"choices": [
{
"message": {
"role": "assistant",
"content": "Hello! As an AI language model, I don't have feelings nor emotions, since I am programmed to assist with language and respond to inquiries. How may I help you today?"
},
"finish_reason": "stop",
"index": 0
}
]
}
I check each request on the page https://platform.openai.com/account/usage
and I noticed that each request/response costs me $0.01. It seemed to me that in accordance with the price on the page https://openai.com/pricing I would pay $0.002 / 1K tokens. Am I misunderstanding something?

[magic moss]

You seem to be calling turbo-0301 model. Did you test with just "gpt-3.5-turbo"? And maybe remove prompt token, completion token and total tokens and just add one field max_tokens and try?

[tawny harbor]

gpt-3.5-turbo is just an alias for the latest snapshot, currently being gpt-3.5-turbo-0301. You can verify this by calling with the gpt-3.5-turbo model and check your usage.