Crawlux ships Whitepaper AEO Scorer, the first AI visibility audit for crypto whitepapers

LONDON · MARCH 12, 2026

Every crypto project has a whitepaper. Most projects assume it works as their definitive reference document for AI engines. The data says otherwise. Across 200 sampled whitepapers in the Crawlux corpus, the median cite rate when AI engines were asked direct questions about the project was 31% across the three major engines. Two thirds of the time, the whitepaper does not get pulled into the answer.

The Scorer runs 23 prompts against any whitepaper URL. The prompts cover the categories AI engines actually get asked: what does the token do, who built it, what is the token supply and distribution, what audits have been done, what jurisdictions can use it, what is the technical architecture, what are the risk disclosures. Three engines run the same prompts in parallel: ChatGPT (GPT-4 class), Perplexity (live search), Claude (Anthropic). The Scorer returns cite rate per engine plus a composite AEO score for the whitepaper.

Why whitepapers underperform the rest of a project's site. Three reasons. First, most whitepapers are PDFs. AI engines extract text from PDFs but lose structure, headings and any schema. The same content rendered as HTML cites 40% more often. Second, IPFS-hosted whitepapers are sometimes blocked or rate-limited by AI crawlers. Hosting on an IPFS-only URL cuts cite rate roughly in half versus a centralized mirror. Third, whitepapers rarely carry the freshness signals (dateModified, version number, changelog link) that AI engines weigh heavily for evolving topics.

The Scorer surfaces all three problems specifically. PDF-hosted whitepapers get flagged with a recommended HTML mirror pattern. IPFS-only whitepapers get a centralized-mirror checklist. Whitepapers without version metadata get a freshness scorecard. Each finding ships with the expected cite-rate lift if remediated, calibrated against the Crawlux audit corpus.

The verticals where this matters most. Token-sale offerings whose whitepaper is the legal document of record. DeFi protocols whose technical depth lives in the whitepaper rather than the marketing site. DAO governance frameworks whose constitution is the whitepaper. RWA tokenization platforms whose disclosure documents double as whitepapers. For each, AI engine miscitation is a real cost: investors get wrong answers about the project from the engines they trust, and the project has no path to correct the answer at scale.

Distribution is unrestricted at crawlux.com/free-tools/whitepaper-aeo-scorer. Each IP gets 10 whitepaper audits per day, no signup. Median runtime: 90 seconds per audit. The Scorer accepts URL or direct PDF upload. Crawlux Pro subscribers get unlimited audits, historical tracking and competitor whitepaper comparisons.

Beyond the score, the tool returns the top 5 highest-impact recommendations specific to the audited whitepaper. Projects that shipped all five recommendations during the beta cohort saw cite rates climb from a median of 28% to 51% within three weeks. The largest single lever in the beta data was adding an HTML mirror of a PDF whitepaper. Median lift from that one change alone: 17 percentage points.

The Scorer joins the free-tools suite at crawlux.com/free-tools and pairs with the AI Citation Checker (released Mar 3) which measures cites for a whole domain rather than a single whitepaper. The pairing covers AEO at both granularities: site-wide visibility and document-specific visibility. Both share methodology with the paid Crawlux Pro audit.

Crawlux is the world's first automated SEO audit tool built for Web3, DeFi and blockchain. The platform runs 23 analyzers across 6 check groups including AI visibility testing across ChatGPT, Perplexity and Claude. Free tier available. Paid tiers from $25 per audit. More at crawlux.com.