Know what you allow.
Interpret crawler-specific policies. Separate search discovery from training preferences and user-requested fetches.
Check crawler policies →Check whether ChatGPT, Claude, Perplexity, Google, Bing and other AI/search systems can discover, understand and access your website.
How AI Crawl Score WorksOne website. An evolving discovery landscape.
Find technical barriers, understand the evidence, and decide what to fix first.
Interpret crawler-specific policies. Separate search discovery from training preferences and user-requested fetches.
Check crawler policies →Inspect raw content, headings, metadata and structured data. Identify what needs deeper rendered-browser review.
Analyze website readiness →Every scored check includes the observation, weight, recommendation, source and limits of what it proves.
Explore the scoring model →No account, installation or access to your analytics. We only request public web pages.
Watch completed scan stages as the engine inspects policies, pages and technical signals.
Open each finding to understand the issue and its limits. Export your report locally for your team.
A readiness score is a diagnostic, not a promise. We do not claim to measure proprietary rankings, guarantee citations, or impersonate verified crawlers.
Training controls are your choice. Missing llms.txt does not lower your score.
AI Crawl Score helps you inspect the technical signals that a public website makes available to search and AI systems. Start with a URL to review crawler policies, page indexability, raw HTML and structured data. The free AI search readiness checker brings those observations into an explainable report. Each finding shows what was tested, the evidence collected and the limits of the conclusion. Use it to plan a technical review, share findings with a developer and check changes after publication.
Your robots.txt file communicates crawling preferences, including different choices for search discovery and training collection. The AI crawler checker compares documented crawler tokens and shows the matched policy for the tested URL. The robots.txt AI checker helps you inspect groups, path rules and sitemap declarations. An allowed rule is evidence of a stated preference, not proof that a crawler visited your site. Server authentication, network conditions and CDN rules can still affect access. Preserve restrictions that protect content you deliberately exclude.
The scanner retrieves public HTML using its own identified client. It checks titles, descriptions, headings, canonical links and indexing directives in that response. This gives you a starting point for understanding content extractability. A page that depends on JavaScript may require a separate rendered browser review. The agent readiness checker also inspects native controls and apparent accessible names. These markup observations help you identify follow-up work, but they cannot confirm keyboard behavior, computed visibility or how an interactive application behaves.
The sitemap health checker discovers a sitemap and inspects XML within a bounded budget. It can help you review structure and host consistency while distinguishing parsed URLs from pages actually tested. The structured data checker examines JSON-LD syntax and declared types in retrieved HTML. Valid JSON alone does not establish complete schema compliance or eligibility for a search feature. Review the specific evidence before making changes, and use the relevant provider documentation when your page needs a more specialized validation.
Use dedicated checks for ChatGPT search readiness, OAI-SearchBot, GPTBot, Claude and Perplexity when you need to investigate one crawler policy. These tools share the same analysis engine and explain their different purposes. The AI robots.txt policy generator creates a snippet to review and merge with your existing rules. The llms.txt generator creates an optional documentation index. Neither generator publishes changes for you. Missing llms.txt or blocking training collection alone does not reduce your search readiness score.
A free scan samples up to five public pages, so its score describes that sample rather than an entire website. Read failed, unknown and untested findings separately. Download the JSON report to keep a local record, make a deliberate change and run another scan when capacity permits. Submitted targets and fetched page bodies are excluded from application logs. Anonymous reports are not saved or shared automatically. Public documentation explains the scoring methodology, data sources and safe scanning boundaries so your team can assess the result.
Each scan requests public resources again; it does not promise continuous monitoring or automatic updates between visits. Crawler source reviews and availability checks support maintenance, while factual changes require review. Accounts, ownership verification, history and alerts remain planned capabilities. The public tools remain available individually, with clear explanations of their scope and limitations. The useful outcome is a clearer technical checklist, with evidence you can inspect and limits you can understand, not a guarantee of rankings, indexing, traffic or citations from any provider.
It samples up to five public pages and checks robots.txt policies, indexing directives, raw HTML, sitemap structure, JSON-LD and basic control naming. Reports explain evidence and untested signals.
The public scanner and text generators are free, with no account or payment required. Scan rate and daily capacity limits apply.
No. The report interprets crawler policies and observations from our generic client. Verified crawler activity requires separate server or CDN evidence.
No. The score is a technical diagnostic. It does not predict or guarantee indexing, rankings, traffic, inclusion or citations.
No. Each scan requests current public resources when you run it. Persistent monitoring, history and alerts are planned and are not enabled.