AI Native · interactive
What an AI feature costs
Every call is billed by tokens in and tokens out, and output tokens cost several times more. Size your feature before launch: users, calls per user, prompt length, answer length and model. Prices are examples per million tokens.
per month
$18,000
1,500,000 requests · $0.0120 each · $1.80 per user
input $9,000output $9,000
Same feature, other models
- Small model ($0.25 in / $1.25 out)$1,500
- Mid model ($3 in / $15 out)$18,000
- Large model ($15 in / $75 out)$90,000
If each user costs $1.80 a month, check that against what they pay. Output is 50% of the bill.
Model
1 Cost = calls × (input tokens × input price + output tokens × output price). System prompts and retrieved context count as input every time.
~4 characters per token in English.
2 3 4
In production
Tools and technologies
Every provider bills by tokens; observability tools track it.
Not tied to one language — use these names when you search docs or design a system.