$9 Pro for api inference and cost

Beginners
1

I am subscribed to the $9 Pro plan and I am using for generating a synth dataset, so a bit of heavy usage here.

/static-proxy?url=https%3A%2F%2Fapi-inference.huggingface.co%2Fmodels%2Fmistralai%2FMixtral-8x7B-Instruct-v0.1%3C%2Fa%3E%3C%2Fp%3E

I just wanted to find out for sure that I am not charged anything else than the montly subscription for using it. This is not a dedicated inference api endpoint, but the normally available one.
There is no indication in the billing section about this usage. I just don’t want surprises.

1 Like
2

I’m also interested in this, as I heavily rely on the Inference API (making 1 request per 10 seconds for 24 hours). I searched the documentation but couldn’t find relevant information.

For reference, here’s the code I use to send requests:

client = AsyncInferenceClient("meta-llama/Meta-Llama-3-8B-Instruct")
chat = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hi, can I reach the moon by jumping?"},
]

response = await client.chat_completion(chat, max_tokens=100, temperature=0.1)
3

I run quite a bit of inference and I was charged only the $9 a month, so it seems…

1 Like