Skip to main content

How AI Visibility Credits Work

Understand how AI visibility credits work, how they’re consumed, and how usage impacts billing.

Updated this week

What it is

Rather than fixed limits, AI Visibility uses credits to reflect real usage. Your credit consumption is determined by three concrete decisions you make: what you track, where you track it, and how often you refresh it.


The three factors that determine credit usage

Every credit used in AI Visibility comes down to a combination of:

  1. Coverage - how many AI engines you track per search term

  2. Cadence - how often those terms are refreshed

  3. Depth - which engines and features you use (e.g. reasoning models, web search, page audits)

Credits scale linearly with these choices. More engines, more frequent updates, or deeper analysis all increase usage.


How refreshes translate into credit usage

Each time you refresh a search term, Keyword.com runs that term against the AI engines you’ve selected.

This creates what we call runs.

A run is best understood as:

Search term × AI engine × refresh event

Examples:

  • 1 search term × 1 engine × 1 refresh → 1 run

  • 1 search term × 3 engines × 1 refresh → 3 runs

  • 10 search terms × 2 engines × 1 refresh → 20 runs

Each run consumes credits based on the engine used.


Refresh cadence options

You control how often refreshes happen per search term:

  • Manual (on demand)

  • Hourly

  • Daily

  • Weekly

More frequent refreshes generate more runs and therefore use more credits.

Tip: Most teams run a small set of critical terms daily, and keep secondary or exploratory terms on weekly or manual refresh.


Credits per AI engine

Different AI engines require different levels of computation. Credit cost reflects this.

AI Engine

Credits per run

Perplexity

0.25

AI Mode

0.25

OpenAI GPT-5o* (auto web search)

0.25

Google AI Overview

1

Gemini

0.25

Claude 4.5

2

DeepSeek V3

1

Perplexity Sonar-Reasoning

1

Perplexity Sonar-Pro

2

Perplexity Sonar-Reasoning-Pro

2

Mistral Large

2

Page Audits

5

xAI Grok (coming soon)

0.25

Bing / Copilot (coming soon)

0.25

* ChatGPT-5o is used most of the time, but the specific model may vary based on availability.


Using the Runway to plan usage

Runway shows how many days of automatic updates you have left at your current settings.

Runway is calculated based on:

  • your remaining credits

  • your active search terms

  • selected engines

  • refresh cadence

A shrinking runway is a signal, not an error. It tells you when your current setup will exhaust credits if nothing changes.

Extending your Runway

You can extend runway without reducing coverage by:

  • switching non-critical terms to weekly or manual refresh

  • reducing the number of engines on exploratory terms

  • pausing low-value terms temporarily


Monthly credit reset & expiry

  • Credits are allocated per billing cycle

  • Unused credits expire at the end of the cycle

  • Credits do not roll over

This encourages active usage planning rather than stockpiling.


What happens when credits run out

If available credits reach zero:

  • Automatic refreshes pause

  • Manual refreshes are blocked

  • You’ll be prompted to top up credits or adjust your plan

There are no automatic overages.

📸 Screenshot placeholder – low credits or blocked refresh state


How teams typically manage credits

Common usage patterns across teams:

  • Core brand or revenue terms → daily refresh, multiple engines

  • Category or competitive terms → weekly refresh

  • Exploratory or experimental terms → manual refresh

  • Page audits → run selectively, not continuously

This approach balances visibility, cost, and signal quality.


Best practices

  • Start broad, then narrow to the engines that matter

  • Use Runway as a planning tool, not a warning

  • Review usage monthly and adjust cadence instead of adding credits blindly

Did this answer your question?