Cloudflare AI neurons usage seems disproportionate
Why it seems to be charging much more neurons than the token usage?
I am using
https://developers.cloudflare.com/workers-ai/models/llama-3.2-3b-instruct/
$0.051 per M input tokens
$0.34 per M output tokens
OK!
Input tokens: 628
Output tokens: 442
So:
0.051 / 10^6 * 656 + 0.34 / 10^6 * 409 = $0.000172516
Since each neuron is $0.011, it should be $0.011 / $0.000172516 = 63.76 neurons
But it shows 873.47 neurons! Why?
I am using
@cf/meta/llama-3.2-3b-instruct. https://developers.cloudflare.com/workers-ai/models/llama-3.2-3b-instruct/
$0.051 per M input tokens
$0.34 per M output tokens
OK!
Input tokens: 628
Output tokens: 442
So:
0.051 / 10^6 * 656 + 0.34 / 10^6 * 409 = $0.000172516
Since each neuron is $0.011, it should be $0.011 / $0.000172516 = 63.76 neurons
But it shows 873.47 neurons! Why?
