Audit and evidence
Public HTML checks, private reports, account recovery codes, saved prompts, manually recorded answers, source analysis, improvement plans, JSON/CSV exports and printable reports.
PRODUCT AVAILABILITY / SEPTEMBER 2026
Coverage you can check.
Choose the right collection method for the question. An engine label, a selected country and an observed local answer are separate facts.
Automation status: awaiting provider activation.
Website audits and manual answer evidence are available. No live automated AI benchmark is published. Each source below has its own collection method.
| AI surface | Manual answer evidence | Automated collection | Cadence |
|---|---|---|---|
| ChatGPT Search | Available; supplied by the user | Not available | No automatic schedule |
| Google AI Mode | Available; supplied by the user | Not available | No automatic schedule |
| Google AI Overviews | Available; supplied by the user | Not available | No automatic schedule |
| Perplexity | Available; supplied by the user | Not available | No automatic schedule |
| Gemini | Available; supplied by the user | Not available | No automatic schedule |
| Microsoft Copilot | Available; supplied by the user | Not available | No automatic schedule |
| Claude | Available; supplied by the user | Not available | No automatic schedule |
| OpenAI Web Search API | Available; supplied by the user | Connector built; activation pending | No automatic schedule |
A chosen market is a requested scope
HowAICite stores country code, region, city, answer language and location provenance. For the OpenAI API, location is an approximate request to the provider. It does not prove where a consumer search ran. Manual observations depend on the location details supplied by their collector.
We do not label a country as measured until a saved observation documents that scope. Cross-country comparisons need equivalent questions, native-language review and enough repeated answers in each cohort.
View 40 country/region options plus a global scope
United States (US) · Canada (CA) · United Kingdom (GB) · Australia (AU) · Vietnam (VN) · Singapore (SG) · India (IN) · Germany (DE) · France (FR) · Spain (ES) · Italy (IT) · Netherlands (NL) · Ireland (IE) · Portugal (PT) · Poland (PL) · Türkiye (TR) · Sweden (SE) · Norway (NO) · Denmark (DK) · Finland (FI) · Switzerland (CH) · Austria (AT) · Belgium (BE) · Brazil (BR) · Mexico (MX) · Argentina (AR) · Colombia (CO) · Japan (JP) · South Korea (KR) · China (CN) · Hong Kong (HK) · Taiwan (TW) · Indonesia (ID) · Thailand (TH) · Malaysia (MY) · Philippines (PH) · United Arab Emirates (AE) · Saudi Arabia (SA) · South Africa (ZA) · New Zealand (NZ) · Global / no country (GLOBAL)
These are scope options, not a statement that every provider supports each location.
View answer-language options
English · Vietnamese · Spanish · French · German · Portuguese · Italian · Japanese · Korean · Chinese (Simplified) · Chinese (Traditional) · Arabic · Hindi · Indonesian · Thai · Malay · Dutch · Turkish · Polish · Other
Answer language is independent of interface language. Public documentation and the operational workspace currently use English.
Product capabilities
Team and distribution
Team invitations, SSO, automated email reports, white-label reports, separate client workspaces and public developer API access.
Recurring measurement
Scheduled multi-engine runs and public cross-market benchmarks. Website audits and planned prompts do not count as AI answer observations.
Review proposed plans and the release log for scope. Availability changes require a documented product release.
Start with your own evidence.
Inspect a website, save an observed answer, and decide what to improve.