LOCAL COST WORKSHEET

API Cost Calculator: Estimate Spend from Tokens and Unit Prices

Enter input tokens, output tokens, and per-million-token prices to calculate API cost locally in the browser. The tool does not query authenticated pricing endpoints or save inputs.

Data processing: Processed locally on this page
Content updated: 2026-07-19

COST FORMULA

Calculate input and output charges separately

Models commonly use different input and output prices. The calculator converts each value from a per-million-token rate before adding a per-request estimate.

01
Input usage
Input tokens / 1,000,000 x input price
02
Output usage
Output tokens / 1,000,000 x output price
03
Per request
Input charge + output charge
04
Batch budget
Per-request total x request count

You provide every price manually, so the result is never presented as live CodeNodex or upstream pricing.

MANUAL INPUT

Enter tokens and unit prices

Use per-million-token prices in one currency. The page does not query model pricing or save these values.

Per-request assumptions

Input tokensOutput tokensInput price / 1M tokensOutput price / 1M tokensExpected request countPrice currency

No network request and no account-price lookup.

LOCAL RESULT

Cost breakdown

This is a text-token estimate. Caching, tiered prices, images, per-tool charges, and exchange rates are outside the formula.

Input cost per request: $0.125
Output cost per request: $0.10
Total per request: $0.225
Estimated total for 1 requests: $0.225

Current formula

(100000 ÷ 1,000,000 × 1.25) + (20000 ÷ 1,000,000 × 5)

How to calculate API token cost

When prices are quoted per million tokens, divide the input-token count by one million and multiply by the input price. Repeat with output tokens and the output price, then multiply the combined result by the expected request count for a batch.

The tool intentionally has no built-in model prices, which prevents stale static values and avoids authenticated pricing requests. Obtain prices in one currency and billing unit from the current model catalog or provider price page first.

Shows input and output charges separately
Supports USD and CNY labels
Uses request count only for batch estimates
Calculates entirely in the current page

Costs that are not included automatically

The basic calculation does not automatically cover cache reads or writes, long-context tiers, images, audio, video, per-tool charges, private discounts, or exchange rates. Treat the output as the text-token portion when those items apply.

Token usage changes with prompts, context, model behavior, and retries. For a monthly budget, calculate both the median and a high percentile from real historical usage.

ESTIMATE BOUNDARY

The estimate may differ from the invoice

Caching, reasoning tokens, tool calls, retries, images, or per-operation items can use separate billing rules that a basic text formula does not cover.

01
Check units
Confirm the price is per million tokens, not per thousand tokens or request
02
Keep one currency
Use one currency for a calculation; the tool does not convert exchange rates
03
Add headroom
Budget for retries, context growth, and task variance
04
Compare billing
Treat provider usage and billing records as final

Use the calculator for budgeting and comparison, not as a replacement for a current price page, channel rules, or the final invoice.

Continue with another tool

Codex Config GeneratorChoose an endpoint, model, and reasoning effort to produce a reviewable Codex CLI configuration; write the real secret securely on your own machine.Open tool Compatibility MatrixInspect client, protocol, version baseline, and verification scope so a text response is not mistaken for complete workflow compatibility.Open tool

NEXT STEP

Move the local result into a real development workflow

Read the matching client guide, then create a separate API key and complete a minimal read-only verification.

View integration guides Sign up and create a key

Enter tokens and unit prices

Use per-million-token prices in one currency. The page does not query model pricing or save these values.

Per-request assumptions

Input tokensOutput tokensInput price / 1M tokensOutput price / 1M tokensExpected request countPrice currency

No network request and no account-price lookup.

Cost breakdown

This is a text-token estimate. Caching, tiered prices, images, per-tool charges, and exchange rates are outside the formula.

Input cost per request

$0.125

Output cost per request

$0.10

Total per request

$0.225

Estimated total for 1 requests

$0.225

Current formula

(100000 ÷ 1,000,000 × 1.25) + (20000 ÷ 1,000,000 × 5)

How to calculate API token cost

Shows input and output charges separately

Supports USD and CNY labels

Uses request count only for batch estimates

Calculates entirely in the current page

Costs that are not included automatically

Token usage changes with prompts, context, model behavior, and retries. For a monthly budget, calculate both the median and a high percentile from real historical usage.

Calculate input and output charges separately

Input usage

Output usage

Per request

Batch budget

Enter tokens and unit prices

Cost breakdown

How to calculate API token cost

Costs that are not included automatically

The estimate may differ from the invoice

Check units

Keep one currency

Add headroom

Compare billing

Continue with another tool

Move the local result into a real development workflow

Calculate input and output charges separately

Input usage

Output usage

Per request

Batch budget

Enter tokens and unit prices

Cost breakdown

How to calculate API token cost

Costs that are not included automatically

The estimate may differ from the invoice

Check units

Keep one currency

Add headroom

Compare billing

Continue with another tool

Move the local result into a real development workflow