Does the calculator include tools, search, grounding or storage fees?

No. The calculator estimates direct text-token charges only. Tool calls, grounding, search, code execution, cache storage, regional uplifts, non-Standard tiers and negotiated discounts are excluded unless represented as token rates in the catalog.

SOURCE-VERIFIED AI API PRICING

AI Model Cost Calculator

Q: Can I compare GPT, Claude and Gemini API costs?

Yes. The optional comparison model uses the same workload and billing date so the two direct token-cost estimates are comparable on the same assumptions.

Q: Does the calculator include cached input pricing?

Yes. Cached input is calculated separately when the selected model has a cached-input rate in the SXF pricing catalog.

Q: How are long-context pricing rules handled?

When the SXF pricing catalog stores a long-context rule for a model, the calculator checks total input tokens and automatically applies the stored multiplier when the threshold is exceeded.

Calculate GPT, Claude and Gemini API token costs using the same source-verified pricing catalog that powers SXF Model Pricing.

WORKLOAD

Build your estimate.

Loading pricing…

Model Compare with optional

Billing date

Request volume

Requests per day

Uncached input tokens Cached input tokens Output tokens

Workload presetsIllustrative starting points — every assumption stays editable.

Presets

Direct text-token charges only. Non-Standard service tiers, tools, grounding, search, code execution, storage, regional uplifts and negotiated pricing are excluded.

ESTIMATED COST

Standard API

Pricing source ↗

PER REQUEST——

PER DAY——

30 DAYS—30-day estimate

1K REQUESTS—same workload

Uncached input—

Cached input—

Output—

Loading pricing dataset…

Same workload and billing date · direct token cost only.

SAME WORKLOAD / ALL CATALOGED MODELS

Compare all models

Sorted by estimated direct token cost only. Capability, latency and output quality are not evaluated.

LOWEST ESTIMATED TOKEN COST — —

Estimated cost of the same workload across all cataloged models
Model	Provider	Per request	30-day cost	Status

CATALOG—tracked models

PROVIDERS—OpenAI · Anthropic · Google

PRICINGStandard APIUSD / 1M tokens

VERIFIED—primary-source catalog

HOW IT WORKS

One workload. Comparable token economics.

The calculator keeps the assumptions visible so GPT, Claude and Gemini estimates use the same token counts, request volume and billing date.

Choose a model

Select a cataloged model from OpenAI, Anthropic or Google. The calculator loads its verified Standard API pricing periods and limits.

Enter token usage

Separate uncached input, cached input and output tokens. Cached input is not blended into ordinary input pricing.

Scale the request

Use requests per day to see per-request, daily, 30-day and 1,000-request estimates from the same workload.

Apply pricing rules

Dated pricing periods and stored long-context multipliers are applied automatically when the catalog defines them.

WHAT IS INCLUDED

Token cost, without hidden assumptions.

This calculator is intentionally scoped to direct token economics so cross-provider estimates remain understandable.

INCLUDED

Direct token charges

Uncached input tokens
Cached input tokens
Output tokens
Stored date-based price schedules
Stored long-context multipliers

EXCLUDED

Provider-specific add-ons

Search, grounding and tool calls
Code execution and storage charges
Batch, Flex, Fast or Priority tiers
Regional or data-residency uplifts
Enterprise or negotiated discounts

CANONICAL DATA SOURCESXF AI Model Pricing & Specs Database

The calculator reads the same JSON catalog used by the Model Pricing database instead of maintaining a second copy of rates.

View model pricing ↗Open JSON dataset

AI MODEL COST FAQ

Common calculator questions.

Use the official provider billing documentation as the final authority before production procurement or budgeting.

How does the AI model cost calculator work?

Choose a model and enter uncached input, cached input and output tokens. SXF selects the stored Standard API pricing period for the billing date, applies any stored long-context rule, and calculates direct token cost per request, per day and for 30 days.

Can I compare GPT, Claude and Gemini API costs?

Yes. Choose a comparison model and both estimates use exactly the same workload and billing date. The comparison is a token-cost comparison only; it does not rank model quality, latency or suitability.

Does the calculator include cached input pricing?

Yes. Cached input has its own field and is priced separately using the cached-input rate stored for the selected model and pricing period.

What does the 30-day estimate mean?

It multiplies the calculated per-request cost by your requests-per-day value and by 30. It is a simple workload projection, not a provider invoice forecast.

Does it include tools, search, grounding or storage fees?

No. Those charges can differ materially across providers and service tiers. The calculator is limited to direct text-token charges represented in the SXF pricing catalog.

How are long-context pricing rules handled?

If a model has a stored long-context rule, the calculator checks total input tokens and applies the cataloged multiplier automatically when the request crosses the stored threshold.