AI AI Prices

AI API pricing glossary

Use this page for support queries like what is input price, cached pricing, batch discount, blend, or budget fit. It defines labels fast, then sends you back to the tool or detail page where the decision happens.

Open compare Open calculator Open FAQ

Use this page like a quick decoder

Glossary is for term definitions, not formulas, trust policy, or final tool selection. Read the term, check where it shows up, then go back to the page where you were making the decision.

Most common labels Input, output, cached, batch, blend

These drive list-page scanning, compare ranking, and calculator interpretation.

Calculator-specific terms Budget fit, cache hit ratio, display currency

Use these when a token price is already being turned into spend or request volume.

Trust and freshness labels Updated, checked, official-first, verified fallback, baseline

These tell you how to interpret source quality and freshness without overclaiming certainty.

Core pricing terms

These definitions show up most often in ranking, compare, and single-model reads.

Input price

The normalized cost for prompt tokens, shown per 1M tokens. Prompt-heavy or retrieval-heavy workloads usually feel this first.

Seen on home and models Use for prompt-heavy requests

Output price

The normalized cost for generated tokens, shown per 1M tokens. Long-form generation often leans on this field more heavily.

Seen on compare and detail Use for generation-heavy requests

Cached pricing

A discounted input rate for the cached share of reusable prompt context. It is not the same thing as batch mode.

Only affects input side Needs provider-listed cache price

Batch discount

A provider-listed reduced execution rate for asynchronous or batch-style processing. Batch counts elsewhere on the site refer to listed batch pricing fields, not blanket provider-wide support.

Separate execution mode Not the same as cache

Blend

A ranking shortcut equal to 75% input price plus 25% output price. It helps scanning, but it is not a substitute for your real workload.

Best for broad list sorting Stop using it once workload is fixed

Budget fit

The maximum monthly request count affordable under the exact request shape and budget shown on the page.

Calculator term Depends on current assumptions

Currency and calculator terms

These matter most once prices are translated into workload cost.

Source currency

The currency attached to the stored provider snapshot. This is the native basis of the captured price.

Stored truth layer

Display currency

The requested viewing currency for homepage, compare, and calculator results. If FX conversion is unavailable, rows fall back to source currency.

Viewing layer only Does not rewrite source snapshot

Cache hit ratio

The share of input tokens assumed to qualify for cached pricing. It only changes the input side of the workload math.

Calculator and compare assumption

1k requests cost

A planning shortcut that turns the current request shape into a comparable per-thousand-requests cost.

Use for quick workload sanity checks

Source and history terms

These explain why a number should be trusted and how pricing history evolves over time.

Updated

The freshest primary signal. It tells you when the latest stored pricing snapshot was captured.

Primary freshness signal

Checked

A secondary verification timestamp for the stored snapshot. It can match the latest snapshot when no separate verification pass exists.

Secondary trust signal

Official-first source

The snapshot came from the active official public pricing source for that provider, which remains the preferred path across the site.

Preferred source path

Verified fallback

The snapshot came from a public fallback source because the preferred official path is not consistently crawlable right now, and that limitation stays visible.

Visible limitation, not hidden substitution

Manual baseline

A manually verified bootstrap reference point used until fresher crawls replace it.

Temporary trust bridge

History window

A 7d, 30d, or 90d comparison against the latest point, only shown when enough stored history exists for that window.

Depends on stored history depth

History confidence

The model detail badge compresses stored history depth into a direct read: warming up, low confidence, or usable signal. It is driven by stored history depth, not guessed from one fresh-looking point.

Shown on model detail first screen
Need a side-by-side decision Return to compare

Use compare after the term is clear and you want to rank a shortlist under one workload setup.

Open compare
Need workload math next Open calculator

Calculator is the right follow-up once terms like cached pricing or budget fit are clear and you need an estimate.

Open calculator
Need why/how answers Open FAQ or methodology

Use FAQ for short support answers and methodology for formulas, history logic, and data flow.

Read methodology