Screaming Frog Review 2026: Free Limits & Licence

The free version is not a trial - it is a permanent tool capped at 500 URLs per crawl, covering the core checks: broken links and errors, redirects, page titles and metadata, robots and directives, hreflang, exact-duplicate detection, site visualisations, and XML sitemap generation. For a small business site of a few hundred pages, that is a complete first audit, and it pairs naturally with the no-cost stack in our free SEO tools guide.

The licence (per the vendor: 'crawl more than 500 URLs and have access to advanced features') unlocks the professional layer - saving and opening crawls, scheduling, crawl comparison, JavaScript rendering, custom robots.txt, custom extraction and custom JavaScript, spelling and grammar checks, near-duplicate detection, forms-based authentication, and segmentation. Two caps remain honest even after paying: crawl size depends on your machine's memory and storage configuration, and licences are per user per year - an agency of five auditors budgets five renewals.

Licensed crawls connect to the measurement stack: Google Analytics, Search Console, and PageSpeed Insights integrations pull performance and search data alongside crawl findings, link-metrics integration brings external authority data in, and a Looker Studio crawl report supports client-facing reporting. That turns a crawl from a list of errors into prioritized work - pages with traffic and problems first - the workflow our Google Analytics alternatives guide complements on the analytics side.

The direct AI integration guide covers OpenAI, Gemini, Anthropic, and Ollama. Remote API costs are separate from the Spider licence and from consumer chatbot subscriptions; limited free API access in some regions is not a blanket free-AI entitlement. Local Ollama does not need the same external API-key setup, but requires suitable local resources. Scope any automated analysis to an approved sample first. Sending page content to a remote model changes the data-handling boundary of what otherwise runs on the desktop.

The published individual licence is US$279, £199, or €245 per user/year, with separate volume bands. US$279 is $23.25 as a monthly equivalent, not an offered monthly subscription. Confirm the relevant currency and tax treatment rather than converting a different listing at an invented exchange rate. A team of auditors needs the applicable number of user licences. Compare that commitment with the specific audit capability already included in SE Ranking or another subscribed tool; a second tool is worthwhile when it removes an actual limitation.

The useful comparison is a complete workflow. Ubersuggest addresses keyword questions, Surfer SEO focuses on content work, and Ahrefs or Semrush provide broader research suites. The Spider can complement those tools, but no verified cost model says every business loses money by not buying it. Our Semrush alternatives guide helps identify the job each purchase is meant to do. Start with the free crawler and upgrade when a named limit or feature blocks useful work.

Against Ahrefs and Semrush: their cloud site-audit modules are convenient, scheduled, and integrated with their research data - but a desktop crawler offers depth the modules do not: full configuration control, custom extraction, JavaScript rendering you configure, local data ownership, and no page-quota pressure from your suite plan. Professionals commonly treat suite audits as monitoring and Spider crawls as investigation.

Against Moz Pro and SE Ranking: both include site audits in affordable suites and suit teams that want one login; neither replaces the configurable depth of a dedicated crawler. Against Mangools: no meaningful overlap - Mangools is keyword-first. The pattern across our SEO category holds: suites overlap each other, the crawler complements them all.

If a crawl finds only one page, first inspect the response status, requested URL, redirects, robots rules, authentication, and links in the returned HTML. A blocked or redirected request is not necessarily a JavaScript problem. On a site you are authorized to inspect, check whether the links to the next pages exist in the raw response and whether the crawler has been restricted to a small scope. Avoid bypassing access controls just to increase the URL count.

The JavaScript SEO guide explains when rendering is useful. Compare raw and rendered titles, canonical declarations, and internal links on a sample before running an expensive full rendered crawl. Look for blocked resources or rendering timeouts that could explain missing content. A server-rendered framework page may already expose everything the audit needs. Record a reproducible URL and the observed response rather than reporting 'the framework is bad for SEO.' Any conclusion about search indexing still needs the relevant search-engine evidence; a local crawl alone cannot supply it.

Export broken-link findings with their source inlinks, group repetitive problems by template, and assign an owner and acceptance check. Missing metadata on one page and a template producing the same missing value on a thousand pages need different fixes. Re-crawl the affected sample after the change rather than calling a ticket complete because a dashboard score improved. Licensed custom extraction can help inspect fields such as an author or modified date, but validate the selector on representative page types before trusting aggregate counts.

The export guide separates CSV, Excel, and Google Sheets outputs from a saved crawl that can be reopened in the application. Saving and reopening are licensed capabilities; licence expiry affects access to saved crawls. Keep portable evidence as well as the application state. Licensed structured-data validation and accessibility audits add checks, not a certification that the entire site meets every policy or user need. Prioritize reproducible issues that affect visitors, then measure whether the fixes had the intended result.

Screaming Frog's SEO Spider earns an 8.8/10 editorial fit score as the standard technical-audit crawler: a permanently useful free tier, a flat $279 per-user annual licence with published volume bands, cross-platform desktop delivery, and integrations that turn crawls into prioritized work. The counterweights are scope-shaped: no keyword or backlink research, per-user annual licensing, machine-dependent crawl capacity, and no cloud collaboration layer. This is a documentation-based review - we have not benchmarked crawl performance ourselves.

Buy the licence when audits are part of someone's job or the 500-URL cap blocks real work; stay free for small sites and occasional checks. Pair it with Ahrefs or Semrush for research - or SE Ranking on a budget - and keep Plausible in view for the measurement side. In a category of rising subscriptions, a flat $279 for the tool most audits start with remains the easiest line item to justify.