HomeCiteBoost

As of October 2026

Feature overview

CiteBoost shows whether AI assistants such as ChatGPT, Gemini, Perplexity and Claude name your brand, which sources they cite and which content you are missing – and helps you create that content. This page describes every feature: what it does, what you enter, what you get and where the numbers come from.

Download as PDF

Where the numbers come from

Measured
Real queries to AI models or real data that CiteBoost reads and counts itself.
AI estimate
An AI (Gemini with Google Search) estimates the value. Useful for orientation, not a measurement from an SEO data provider.
Own calculation
A CiteBoost metric derived from measured data; the formula is explained in the app.
Your data
Data from your own accounts, e.g. Google Search Console.

Content Studio

From finding to finished article: Content Studio reads your website, finds missing topics and questions where AI assistants do not name your brand, and turns them into researched briefs and articles – with sources, FAQ and images if you like.

1. Website analysis

  • Reads your website via its sitemaps (from robots.txt, /sitemap.xml and sitemap indexes, also compressed), otherwise via links. robots.txt is respected and only your own domain is read.
  • Per page: title, meta description, H1 and subheadings, word count of the main content, structured data, language, canonical, noindex, internal links and last modified date.
  • Research with Google Search puts your topic in context. Result: summary, target audience, topic clusters (strong, medium, weak) and 12 to 20 concrete opportunities.
  • Prompts with low AI visibility add further opportunities; thin pages (under 300 words) are flagged.
  • The analysis runs in the background; you see the progress.

2. Opportunities

  • Types: new content, expand existing content, FAQ, refresh.
  • Per opportunity: title, keyword, search intent, a rating from 1 to 100 (AI assessment), reasoning and affected URLs; for AI gaps also the prompt, the measured visibility and the cited domains.
  • Sort, filter, dismiss and restore.

3. Content brief

  • Research with Google Search; sources come exclusively from the search results.
  • Contents: H1, keywords, search intent, audience, goal, angle, format, length, tone, outline with H2/H3, must-answer questions, entities, internal links (only to pages that exist on your website), insights from the top results, AI visibility notes, image ideas, meta data, call to action and sources.
  • Extra care for sensitive topics such as health, finance and law.
  • Every field is editable; an article is created only from an approved brief. Markdown export.

4. Article

  • Written section by section, with context from the previous section and a word target per section. Form of address and language are detected automatically.
  • No invented facts; statements are backed by the brief’s sources and linked.
  • Meta title, meta description and 3 to 6 FAQs as a visible section.
  • Edit the heading, meta data and sections; rewrite individual sections with your own instruction.
  • Export: copy Markdown or HTML, .md and .html files (with meta data, preview image and FAQPage schema) or a ZIP with all images.

5. Images

  • Cover image and one image per section in 16:9, 4:3, 3:2 or 1:1; the image prompt is suggested and editable.
  • Generation rules: no text in the image, no real people or brands.
  • Upload your own images (JPG, PNG, WebP): they are resized and metadata such as GPS is removed. Alt text is required.
  • Labelling: optional “AI” badge and caption templates; AI images carry the image model’s signed C2PA provenance certificate.

Limitations

  • The website analysis contains no search volumes or rankings.
  • No direct export to a CMS such as WordPress – you take over the files.
  • Existing pages are not rewritten automatically; “expand” creates a new text based on the brief.

AI visibility (GEO)

Generative Engine Optimization: how often and how are you named in answers from AI assistants? CiteBoost asks AI models real questions and analyzes the answers.

AI Projects

Brand, competitors and languages in one place.

Your data
What it does
A project holds everything CiteBoost needs to know about a brand. Prompts, reports, Citation Builder and Content Studio all use it.
Input
Project name, brand, industry, website, competitors (as names or domains) and the languages you want to monitor.
Output
  • One project per brand or client
  • Competitors that are counted automatically in every AI answer
Data source
Your input.
Typical uses
Keep several brands or clients separate · Define the competitors to compare against

Prompt Library

The questions your customers ask AI assistants – measured.

Measured
What it does
You create questions (prompts) your audience might ask an AI assistant, e.g. “Which software is suitable for …?”. Each prompt runs as a real query against the selected AI models. CiteBoost analyzes every answer and stores it in full.
Input
Prompt text, brand, industry, category, tags, priority, the AI models and optionally a schedule (daily, weekly, monthly). An AI suggests suitable prompts based on brand, industry and website.
Output
  • Visibility: share of models that name your brand
  • List position of the mention (in numbered or bulleted answers)
  • Tone of each mention (positive, neutral, negative – rated by an AI)
  • Mention share of your competitors in the same answers
  • Cited sources and the full answer texts
  • History per prompt with a comparison of two runs
Data source
Measured: real queries to the models through their APIs (via OpenRouter). The brand is detected in the answer text (all spellings, with word boundaries).
Typical uses
The basis of any AI visibility monitoring · Automatic runs on a schedule, followed by evaluation in the Alert Center · Track the effect of new content over time
Limitations
  • Models are queried through their APIs, not the apps with your account – without personalisation or chat history.
  • Perplexity models always search the web live; for other models web search can be enabled where available.
  • AI answers vary between runs. Several prompts and runs give a reliable picture.

AI Overview

All AI metrics at a glance.

Measured
What it does
Summarizes your prompt results for a period and compares your brand with your competitors – counted in the same AI answers.
Input
Project and period (7, 30, 90 or 365 days).
Output
  • Average visibility, tone, position and number of mentions
  • Trend over the period
  • Brand comparison with your competitors
  • Most cited sources and the latest answers
  • PDF and CSV export
Data source
Evaluation of your stored measurements.
Typical uses
Weekly overview · Reporting to clients or management

AI Sources

Which websites AI models cite.

MeasuredOwn calculation
What it does
Shows which domains and URLs AI models cite as sources in answers to your prompts.
Input
Project (optional).
Output
  • Top domains and URLs with source type (editorial, community, reference, news, company)
  • Frequency and trend (last 30 days vs. the 30 days before)
  • Source index: own metric from frequency, source type and citations per URL
  • CSV and PDF export
Data source
Measured answers; the source index is an own calculation, not an external domain authority.
Typical uses
Find out where you should be present · Basis for PR and content planning
Limitations
  • Only sources that models actually name in their answers are counted.

Live Monitor

The quick spot check.

Measured
What it does
Immediately asks the selected AI models which providers or tools they recommend for a topic and shows whether and where your brand is named.
Input
Topic or product category, brand, optional website, AI models.
Output
  • Per model the answer, mention, list position and context
  • A score from mention rate and list position
Data source
Measured (real queries).
Typical uses
Spontaneous checks, e.g. before presentations or product launches
Limitations
  • Fixed question per check; the Prompt Library is meant for regular, comparable measurements.

Alert Center

Notifications about real changes.

Measured
What it does
After every prompt run CiteBoost compares your brand’s new standing with the previous one and notifies you when something changed beyond your thresholds.
Input
Rules with type (visibility dropped or rose, competitor ahead, tone changed, position changed, new source), threshold in percent or points, brands, competitors and channels.
Output
  • Alerts in the dashboard, by email or via webhook
  • Delivery immediately or as a daily or weekly digest
  • Webhook messages that Slack and Discord display directly; JSON for Zapier, Make or n8n
Data source
Comparison of measured standings: per brand the latest result per prompt of the last 30 days counts.
Typical uses
Together with the Prompt Library schedule: automatic monitoring

Multi-Language Tracking

Your visibility in up to 15 languages.

Measured
What it does
Translates your prompts into the selected languages and asks the AI models in each language. You see in which languages your brand is named.
Input
Languages (German, English, French, Spanish, Italian, Portuguese, Dutch, Polish, Swedish, Danish, Norwegian, Finnish, Japanese, Chinese, Korean), prompts and AI models.
Output
  • Visibility and mentions per language, also per model
  • Comparison with the previous language run
  • Prompt × language matrix, CSV export
Data source
Measured with translated prompts. An AI creates the translations; you can review and edit each one.
Typical uses
International brands · Before entering new markets
Limitations
  • The brand is detected as written; transliterations (e.g. Katakana) are not detected.

Citation Builder

Where you should be mentioned – and whether your pages are citable.

MeasuredOwn calculation
What it does
Shows the domains AI models cite on your topics while your brand is missing from the same answer. A second part checks, without AI, whether one of your pages can be read and cited by AI crawlers.
Input
Brand and period; a URL for the check.
Output
  • Source gaps: cited domains with answers without your brand, the competitors named there, models, prompts and URLs
  • Competitor websites are detected and marked separately
  • Status and notes per domain, CSV export
  • Citability check with 16 checkpoints, e.g. “content visible without JavaScript”, robots.txt access for GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, ClaudeBot, Google-Extended and CCBot, structure, structured data, FAQ, author and date, source links, llms.txt
Data source
Measured answers; the check reads the page like a crawler (without JavaScript). The citability index is an own calculation and does not predict whether an AI will actually cite the page.
Typical uses
PR and outreach planning · Check your own pages before revising them
Limitations
  • The approach tips per source type are general.

SEO tools

Tools for classic search. Where numbers such as search volume or difficulty appear, they are AI estimates – CiteBoost is not connected to an SEO data provider. All results can be exported as PDF with your own branding (except Search Analytics).

Page Audit

Check a page for structure, content and technical basics.

MeasuredAI estimate
What it does
Loads the page as delivered without JavaScript and has an AI assess it: title, headings, meta description, structured data and content quality.
Input
One URL.
Output
  • Overall score and sub-scores for technical, content and authority
  • Short company summary and detected keywords
  • Issues by severity with a suggested fix
Data source
The page is read; ratings and scores are an AI assessment.
Typical uses
Check important pages before and after publishing
Limitations
  • One page, not the whole website; no load-time measurement.

Keyword Explorer

Keyword ideas and search intent.

AI estimate
What it does
Researches related search terms for a seed keyword and analyzes the current top results.
Input
Seed keyword and number of ideas.
Output
  • Ideas as questions, phrase match and related terms
  • Search intent, estimated search volume, difficulty and CPC
  • Overview of the top results and ranking chance
Data source
AI estimate (Gemini with Google Search).
Typical uses
Topic and content planning

Ranking Check

Where does a domain appear?

AI estimate
What it does
Checks where a domain appears for a keyword in results retrieved via Google Search and shows the top 10. In “Discover” mode the AI estimates which keywords a domain probably ranks for.
Input
Domain and keyword, or domain only (Discover).
Output
  • Found position or estimated range
  • Top 10 results
  • Keyword list for a domain (Discover)
Data source
Retrieved via Google Search by the AI; if the domain is not found, the position is an estimate.
Typical uses
Spot checks after content updates
Limitations
  • Snapshot without location or personalisation, no daily tracking.

Competitor Analysis

Two domains compared.

MeasuredAI estimate
What it does
Reads the homepages of two domains and has an AI assess the differences.
Input
Your domain and a competitor’s domain.
Output
  • Title, meta description, headings, images, links, schema and canonical of both homepages
  • Strengths, weaknesses and opportunity gaps
  • Estimated authority and traffic level
Data source
On-page facts measured; ratings, authority and traffic are AI estimates.
Typical uses
Strategy workshops · Before campaigns

Content Explorer

Successful content on a topic.

AI estimate
What it does
Researches successful content and content ideas on a topic.
Input
Topic and amount.
Output
  • Example content with format, approximate length and why it works
  • Takeaways and recommendations
Data source
AI research with Google Search; word counts and engagement are estimates.
Typical uses
Ideation
Limitations
  • No share, backlink or traffic data.

Keyword Gap

Topics a competitor covers – and you do not.

AI estimate
What it does
Compares your domain with a competitor and estimates which keywords they cover and you do not.
Input
Your domain and a competitor’s domain.
Output
  • Gap list with estimated volume, difficulty, priority and example URL
  • Quick wins, strengths of both sides and content recommendations
Data source
AI estimate with Google Search, no real ranking data.
Typical uses
Content planning after competitor analyses

Search Analytics

Your Google Search Console data in the app.

Your data
What it does
Connects your Google account (read-only) and shows the data of your Search Console properties.
Input
Google account, property and period (7, 28 or 90 days).
Output
  • Top queries with clicks, impressions, CTR and average position
Data source
Real data from Google Search Console. Credentials are stored encrypted and revoked when you disconnect.
Typical uses
Performance reviews · Spot potential for new content

Platform

AI models
Queries run via OpenRouter with models from OpenAI, Anthropic, Google, Perplexity, xAI and others. You choose the models per run; analyses and texts in Content Studio are created by Google Gemini.
API access and MCP
With your own API key you can use CiteBoost from your own scripts (REST API) or directly from AI assistants with MCP support such as Claude or Cursor – with tools for prompts, history, AI overview, source gaps, citability check and SEO tools, among others. Keys are stored only as a hash.
Exports
PDF reports with your own branding (company name, report title, client) and CSV exports in AI Overview, AI Sources, Prompt Library, Multi-Language and Citation Builder.
Free AI check
On the homepage you can check without signing up whether an AI model knows and recommends your brand – one model, one question, no web search.
Privacy and security
Details on hosting, data processing and security are available in the Trust Center.

Try it yourself

Start with the free plan or test your brand directly on the homepage.

Go to CiteBoost