AI surfaces

Measured where shoppers ask.

rankyn measures the surfaces your shoppers actually ask, from ChatGPT to Google AI Overview. Each surface is measured the way it really works.

Get a free auditSee pricing
  • ChatGPT
  • AI Overview
  • AI Mode
  • Gemini
  • Perplexity
  • ChatGPT (API)Enterprise
  • Claude (API)Enterprise

The AI surfaces

rankyn measures a pool of 5 AI surfaces, each the way it really works, plus ChatGPT (API) and Claude (API) for Enterprise. The table shows how many people use each one, with source and date, and how rankyn measures it.

The AI surfaces
SurfaceUsersHow rankyn measures it
ChatGPT (App)1.2B+ weekly usersOpenAI · Oct 2026Real interfaceMeasured in the ChatGPT interface shoppers use, set to your market.
Google AI Overview2.5B+ monthly active usersGoogle · May 2026Real interfaceRead from real Google results pages in your market and language.
Google AI Mode1B+ monthly active usersGoogle · Jul 2026Real interfaceMeasured in the real AI Mode interface, set to your market.
Gemini (App)1B+ monthly usersGoogle · Aug 2026Real interfaceMeasured in the real Gemini interface, set to your market.
Perplexity (API)780M queries a monthPerplexity CEO Aravind Srinivas · May 2025APIMeasured through the Perplexity API, with web search on every call.
ChatGPT (API)Enterprise1.2B+ weekly ChatGPT usersOpenAI · Oct 2026APIMeasured through the OpenAI API with web search, the layer apps and agents build on.
Claude (API)Enterprise~56M monthly app users (est.)Sensor Tower estimate · Q2 2026APIMeasured through the Anthropic API with web search.

Why more than one surface

One surface never tells the whole story. Shoppers ask different assistants, and the same product can be recommended in one and left out in another.

  • Your buyers spread out

    Your customers ask ChatGPT, Google, and other assistants. Measuring only one leaves most of your answers unseen.

  • Each surface ranks differently

    The same prompt can recommend you on one surface and a competitor on another. Several surfaces show the real pattern.

  • One picture, not one snapshot

    rankyn brings the surfaces together into a recommendation rate per surface, so you act on the whole market instead of a single answer.

“Best magnesium for better sleep?”
SurfaceRecommended firstYour productHalvor & CoLumae LabsOstrel
ChatGPT
AI Overview
AI Mode
Gemini
Perplexity
ClaudeEnterprise

2 of 6surfaces recommend your product first. The others each name someone else.

Measured the way it really works

rankyn names the method for every surface: the real interface where shoppers type, or the provider’s API with web search on. You always know which one you are looking at.

Real interface

Where shoppers type

Where shoppers use an interface, rankyn measures that interface in your market and language, not a copy of it.

Measured this way

  • ChatGPT
  • AI Overview
  • AI Mode
  • Gemini

API

What apps and agents build on

Perplexity, and ChatGPT and Claude for Enterprise, are measured through the provider’s API with web search on: the layer that apps and agents build on.

Measured this way

  • Perplexity
  • ChatGPT (API)Enterprise
  • Claude (API)Enterprise

We build what you need.

Need another AI surface?

Tell us which surface you need and the team will look into it. Every message is read and it shapes the roadmap.

Leave your email if you’d like a reply.

Measure every surface that matters.

See where your shoppers ask, and who wins there. Start with a free audit measured on the real AI surfaces.

Get a free auditSee pricing

Questions about AI surfaces

Which surfaces rankyn measures, how each one is measured, and how yours are chosen. Can’t find your answer? Write to hello@rankyn.ai.

Which surfaces does rankyn measure?

rankyn measures a pool of 5 surfaces your shoppers actually ask, from ChatGPT to Google AI Overview. Enterprise adds ChatGPT (API) and Claude (API).

Why measure several surfaces?

rankyn measures several surfaces because your buyers spread out across assistants. One assistant never stands in for the whole market, and the same product can win on one surface and lose on another.

How does rankyn measure a surface?

rankyn measures each surface the way people reach it: in the real interface where shoppers type, set to your market, or through the provider’s API with web search on. The method is named for every surface.

What is the difference between interface and API?

An interface is where a shopper types, and an API is what apps and agents build on. rankyn measures each surface on its own terms and names the method, so you know what you are looking at.

How are my surfaces chosen?

You pick your surfaces in rankyn from the pool, up to the number your plan includes. rankyn suggests a starting set that you can change, so you measure where your buyers actually ask.