PayAPIKey

AI Model Cost Calculator

Compare editable AI API price presets and estimate token costs, cache savings, monthly usage, gateway margin, and sustainable pricing.

Inputs

Editable usage, rate, and operating assumptions

Outputs

Scenario totals and unit economics

Best for

Compare official API model price structures

AI Model Cost & Gateway Pricing Calculator

Estimate official API token costs, cache savings, delivery costs, and sustainable pricing for a compliant AI product.

Selecting a preset updates the editable token rates below.

OpenAI GPT-5.6 Sol · gpt-5.6-sol · GA · verified 2026-08-03

Long-context pricing applies to the whole request when average input tokens are more than 272,000. Crossing the boundary reloads the official tier; other edits stay manual.

Published cache charges — not included in this estimateCache write: $6.25/1M

Official standard and batch token rates; unpublished price components remain absent. Rates above 272k prompt tokens use the long-context tier.

Monthly requests
300,000
Across the chosen billing period
Upstream API cost
$7,893
Tokens, add-ons, and retries
Gateway net margin
-69.4%
At planned monthly revenue
Recommended revenue
$14,853
$49.51 per 1,000 requests
ItemBasisValue
Uncached input273M tokens at $5.00/1M$1,365
Cached input147M tokens at $0.50/1M$73.50
Output tokens210M tokens at $30.00/1M$6,300
Tool and media add-ons300,000 requests$0.00
Retry cost2% of base upstream cost$155
Total direct costUpstream API plus monthly platform cost$8,243
Recommended pricePer 1M blended tokens$23.58
Planned net profit4.5% revenue deductions-$3,468.27
At these assumptions, cache hits save about $662. To cover costs, revenue deductions, and a 40% target net margin, monthly revenue should be at least $14,853.

Reference presets verified 2026-08-03. Prices change frequently. Always verify the linked official provider page before making billing decisions.

Cache writes, hourly cache storage, regional processing, enterprise terms, taxes, and other provider-specific charges are disclosed where known but are not silently added. Enter any applicable amount through the editable rates or manual add-on fields.

Pricing methodologyMachine-readable JSON

For official API pricing, BYOK products, team planning, and compliant gateway economics. Not for account sharing, credential resale, or proxy access.

Frequently asked questions

Practical answers for applying this calculator to a production API billing or usage plan.

Are the AI model prices kept up to date automatically?

No. The calculator starts with a small set of reference presets, but every price field is editable. Providers change pricing, model availability, cache rules, and tool fees, so verify the current official provider page before using the result in a billing decision.

How does cache hit rate affect AI API cost?

The calculator treats the selected percentage of input tokens as cached and prices those tokens at the cached-input rate. A higher cache hit rate lowers cost only when the provider offers a discounted cache-read price and your requests reuse eligible prompt content.

Does the recommended revenue include payment fees and refunds?

Yes. Payment fees, expected refunds, and bad debt are modeled as percentages of revenue. They are included with target margin when calculating the monthly revenue needed to cover direct delivery and platform costs.

Can I use this for account sharing or subscription proxy access?

No. It is designed for official API pricing, BYOK products, legitimate team planning, and compliant AI gateway economics. It is not for account sharing, credential resale, or proxy access.