grok-4-1-fast-reasoning - overcharge

Hi there,

I'm reaching out to see if someone can provide some clarification or guidance regarding the pricing for grok-4-1-fast-reasoning.

I was using grok-4-1-fast-reasoning to run some experiments. According to the pricing documentation: https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/

the model is priced at $0.20 per million input tokens and $0.50 per million output tokens.

My usage metrics show:

Total token count: 1.29M (9,548 average per request)

Input tokens: 1.25M (9,285 average per request)

Output tokens: 35.52K

Estimated total cost: $9.42

Based on the documented pricing, I would have expected the estimated cost to be significantly lower than what is being reported in the metrics. The reported estimate of $9.42 doesn't seem to align with the published input and output token rates.

Could someone help me understand how the estimated cost is calculated? Are there any additional charges, different pricing tiers, or other factors that could explain the discrepancy?

Thanks in advance for any clarification.