Customer support
10,000 requests per month; one call per request; 1,400 input words and 300 output words per call. At the calculator’s editable estimate of 1.4 tokens per word, that is 1,960 input and 420 output tokens.
Edit the support example →KINGY AI / PRICE INDEX COMPANION
Start with the work, then count every billed call.
← AI Price Index · October 2026 source edition
The monthly index describes a fixed basket at a stated date. Your bill depends on a particular model, token mix and number of calls. The existing calculator uses separately maintained current pricing evidence; its dates can be newer than the October edition.
10,000 requests per month; one call per request; 1,400 input words and 300 output words per call. At the calculator’s editable estimate of 1.4 tokens per word, that is 1,960 input and 420 output tokens.
Edit the support example →1,000 documents per month; one call per document; 1,500 input and 200 output words. The same conversion gives 2,100 input and 280 output tokens. Large documents, image processing and retries may need additional allowances.
Edit the extraction example →500 tasks per month; three calls per task; 6,000 input and 2,000 output words per call. This means 1,500 calls, each estimated at 8,400 input and 2,800 output tokens. Count repeated context and unsuccessful attempts too.
Edit the coding example →Monthly API cost = calls × (input tokens × input rate + output tokens × output rate) ÷ 1,000,000, when both rates are per million tokens. Add other billed charges separately.
For example, the October 19 snapshot records GPT-6.1 Sol at $2 input and $10 output per million tokens. The support example therefore gives 10,000 × (1,960 × $2 + 420 × $10) ÷ 1,000,000 = $81.20. This is a frozen October illustration, not a fresh price quote. Open the calculator to inspect current evidence and edit the assumptions.
Expired or held source evidence should remain unavailable in the calculator. A monthly edition must never silently reset its current-price verification date. Subscription fees are separate from API billing and do not imply a token allowance.