Screaming Frog SEO Spider
vs
lychee

Screaming Frog vs lychee: Crawl a Site or Check Its Files

Our recommendation

I'd choose Screaming Frog for a public website's source-page repair report, and lychee for Markdown or built HTML checked during development. Screaming Frog vs lychee starts with discovery: one is a website crawler; the other checks links in the inputs you give it. A homepage alone isn't a recursive lychee audit.

Two crawler symbols inspect a paper page network with one broken link.
On this page

Screaming Frog vs lychee at a glance #

FactorScreaming Frog SEO Spiderlychee
Free priceFree editionFree open-source application
Paid optionAnnual license for larger crawls and advanced featuresNo paid license required
Free limit500 URLs per crawlNo license-based URL allowance
PlatformsWindows, macOS, LinuxWindows, macOS, Linux binaries
DiscoveryFollows website links during a crawlReads specified files, directories or remote pages; no recursive remote-site crawl
Input formatsWebsite URLs and URL listsMarkdown, HTML, text and remote inputs
ReportsDesktop tabs, Inlinks and bulk exportsTerminal and structured output, including JSON
AutomationScheduling and advanced configuration in paid editionCLI and GitHub Action
JavaScriptRendering available with paid licenseLink extraction and HTTP checks; no browser rendering
Open sourceCommercial closed-source applicationMIT/Apache-2.0 project
MaintenanceCurrent 24.0 release seriesActive release history and maintained docs
Processing and privacyLocal crawler; requests visit checked sitesLocal or build-runner process; online checks make requests, offline mode blocks them

Sources: Screaming Frog features, 24.0 release, lychee repository, usage and how it works.

Site discovery: Screaming Frog finds the pages for you #

Start Screaming Frog at a website address, and its crawl discovers linked URLs. For a public company site, that is the convenient starting point. You don't need a local copy of every page or a prepared file list before seeing where the links go.

lychee's official usage guide explicitly says a remote URL checks links on that page without recursive website crawling. It can traverse a local directory of supported files, or accept multiple inputs. Those are different ways to define the scan's scope.

I'd use Screaming Frog when an editor says an old link is broken but can't identify every page containing it. The crawler's discovery and report help locate the work. I'd use lychee when the exact documentation files are already part of a repository and the publishing process can supply them.

Consider a blog with a hundred posts. A desktop crawl begins with the live links. A source-file check begins with the files containing those posts or the generated HTML. Both can help, but they see the content supplied through their respective workflows. An unlinked or omitted page still needs a way into the check.

Winner for discovering a public site's linked pages: Screaming Frog.

Before publication: lychee can inspect the build you intend to upload #

The strongest reason to choose lychee is timing. Check your built files before publishing them, and you can catch a reference to a file the new build doesn't contain. A live-site crawl sees the currently deployed website, which may still have the old asset.

This is where configuration matters. In issue 1718, opened May 2025, a user struggled with --base-url, --root-dir and remapping while checking a built folder. Contributors worked through the interactions, and a local-folder recipe was added to the documentation in January 2026. The discussion was closed after that documentation work.

The current local-folder recipe explains how to map future public URLs to the files on your computer. That's the detail I'd follow before putting the checker into a release process. Without the right mapping, a fully qualified link can send the check to today's public site instead of the new build.

Screaming Frog can audit a reachable staging website, which is a straightforward route when a team already has staging. lychee makes more sense when the build is a folder of HTML waiting for upload and the same automated check should run each time.

Winner for checking source docs and a local build before release: lychee.

Repair reports: a content editor needs source and destination together #

Screaming Frog's broken-link tutorial shows the Inlinks report for a selected failed URL. I'd use that when assigning repairs to someone working in a CMS. The page containing the reference matters as much as the failed destination.

lychee reports results against its inputs and supports machine-readable output. The project is built for use in repositories and automated workflows. A failed reference in a Markdown file can be a clear developer task without opening a desktop audit application.

Choose the report for the person doing the correction. A writer maintaining a company blog may want a spreadsheet with source page, link text and failed URL. A developer reviewing a documentation change may want the build job to point at the affected file. Neither person needs a hundred unrelated SEO warnings to fix one typo.

The Screaming Frog guide covers the source-page export. For other browser reporting options, Dr. Link Check and Dead Link Checker appear in the free broken-link checker list.

Winner for a CMS editor's repair spreadsheet: Screaming Frog; for repository feedback: lychee.

Repeated checks: pay for desktop features or configure a CLI #

Screaming Frog's free version is useful for a small manual audit, but its feature comparison places scheduling, saving and opening crawls, and advanced configuration in the paid edition. Larger crawls also need the annual license.

lychee is a free CLI tool with an official GitHub Action. The getting-started documentation describes binaries for Windows, Mac and Linux and the action route. I'd consider it when link checking should be part of each documentation change, with configuration stored beside the content.

Free software still requires decisions. Which directories count? Which external domains should be checked? How will the team inspect a failure? A poorly chosen check can interrupt work with irrelevant reports, or silently omit the pages that matter.

I'd keep the first automated check small enough to understand. Start with one documentation folder, review its results and expand once the input and path rules are right. A transparent check that the team trusts is more valuable than a broad job everyone learns to ignore.

Winner for free repository automation: lychee; for a scheduled desktop audit with visual configuration: paid Screaming Frog.

What users say about the effort involved #

In a December 2025 discussion, SEO practitioners strongly praised Screaming Frog's value for professional audits. A reply suggested trying the free edition first, while another made the site's size part of the paid decision. The thread also shows beginners asking what to do with the crawl results.

I'd take that last point seriously when choosing a tool for a small business owner. A detailed report needs someone who understands which finding deserves a change. Buying the larger allowance doesn't teach the editor how to choose a replacement for a deleted resource.

lychee's local-folder discussion shows a different learning cost. A contributor provided careful explanations of the path flags, and the user thanked them while arguing that checking a folder should be easier. Maintainers discussed making the behavior understandable without adding options they couldn't maintain. The resulting recipe helps, but the setup remains more technical than entering a site address.

For an owner making occasional content edits, I'd choose the clearer repair workflow. For a developer already maintaining builds and repository checks, lychee's configuration belongs naturally with the other project settings. Match the effort to the work you'll repeat.

Choose Screaming Frog if… #

  • You want to discover linked pages on a live or staging website.
  • The person making repairs prefers a visual report and source-page export.
  • The free 500-URL crawl is enough, or you have a reason to buy the larger allowance.
  • Broken links are part of a wider desktop SEO audit.

Choose lychee if… #

  • The content lives in Markdown or generated HTML files.
  • You want checks during repository changes or before deployment.
  • You're willing to define the inputs and local URL mappings.
  • A CLI result or structured report fits the team's workflow.

Dr. Link Check is my browser alternative for occasional scans. Its free plan has two projects and 1,500 unique links per project. Recurring scheduling requires a paid plan.

Ahrefs Webmaster Tools works for verified websites, with 5,000 monthly Site Audit crawl credits per project. It is attractive when the owner also wants related SEO data, though ownership verification adds setup.

SiteOne Crawler is a free open-source option for wider local technical reports. Its variety of audit and export modes asks more of a first-time user than a dedicated link form. See Screaming Frog alternatives for more choices.

FAQ #

Can lychee crawl my entire website from the homepage?

No. Its documented remote-URL workflow checks links on the supplied page without recursive site crawling. Give it the required page inputs or the built local files. I'd choose a website crawler when automatic page discovery is the main requirement. Local directory traversal is a separate lychee feature.

Yes, its free edition includes broken-link checking within a 500-URL crawl. The annual paid license raises the crawl allowance and provides features such as JavaScript rendering and scheduling. Start with the free report for a small manual task, then decide whether an upgrade solves an actual restriction in your workflow.

Does lychee upload my documentation files to a service?

The CLI reads inputs in the environment where you run it; it isn't a hosted document-upload service. Online checks still request the destinations being checked. Its offline mode blocks network requests. Choose that mode for local-file checking, and configure the path mappings for the build you actually intend to validate.

Who maintains lychee?

lychee is maintained through the lycheeverse open-source project on GitHub. Its repository includes the source, issue discussions and release history, while the official docs explain usage and configuration. It is separate from other products named Lychee. The CLI product page links to the correct project.

Sources #

Official docs and feature links appear beside their claims. Community and developer discussion allocated to this page: Screaming Frog value discussion and lychee local-folder issue 1718.

Featured illustration: WebIndexly, created with OpenAI image generation.

Directory details
Directory details: Screaming Frog SEO Spider and lychee
Directory detailScreaming Frog SEO Spiderlychee
What it is
Crawl websites for SEO audits on Windows, macOS and Ubuntu using the Screaming Frog SEO Spider.
lychee checks links in documents and websites for broken hyperlinks and email addresses.
Website
Type
Website
App
Topic
Pricing
Paid
Not confirmed
Platforms
Linux, macOS, Web, Windows
Windows, macOS, Linux
Open source
Not stated
Yes
Language
English
English
Alternatives
C
Chris McClureCo-founder & Editor, WebIndexly

Chris McClure is a tech blogger and co-founder of WebIndexly, with a strong interest in SEO, web architecture, user experience and the technologies that shape the modern web. He focuses on making complex technical topics easier to understand and more useful for developers, creators and businesses…