Back in April 2026, I spent a few weeks exploring SitecoreAI Pathway and presented my findings at the Sitecore User Group Coimbatore (SUGCBE). I wrote three posts about it - a high-level comparison with the old migration tool, a step-by-step walkthrough of the Sitecore Website path, and a walkthrough of the Any Website path.
In those posts, I called out the beta limitations - 50 URL cap, single language per run, no image handling for crawled sites, no way to guide the AI's decisions.
I recently went through the latest official documentation and several of those limitations are gone. Here is what I found.
The 50 URL Limit Is Now 200
When I crawled my blog with the Any Website path, Pathway picked up 50 pages from the sitemap and stopped. My blog has more than 50 posts, so I was leaving content behind.
The docs now say the crawler supports up to 200 URLs. For a blog like mine, 200 covers everything. For larger enterprise sites you would still need multiple runs, but it is a much more practical starting point.
Multi-Language Support
In my earlier testing, Pathway supported only a single language per migration run. If your site had content in English, French, and German, you needed three separate runs.
The docs now say "SitecoreAI Pathway supports sites in multiple languages." I have not tested this hands-on yet, so I cannot say how language detection and mapping work in practice. But the single-language constraint was one of the first things people asked me about after my SUGCBE talk, so it is good to see it addressed.
Image Import for the Any Website Path
When I used the Any Website path to crawl nehemiahj.com, Pathway migrated the page content but not the images. For the Sitecore Website path, you could use the old XM to XM Cloud Migration Tool to move media separately. But for the Any Website path, there was no image solution at all.
The docs now say: "If enabled, the SitecoreAI Pathway app also imports images from the scraped website." And for non-Sitecore sites, "media is scraped from the source site and moved to /sitecore/media library/project."
So the Any Website path can now handle both content and images in a single run. The "if enabled" part suggests there is a toggle for this - I want to find it and see how the imported images look in the media library.
Specific URL List Instead of Just Sitemap
During my April testing, the crawler relied entirely on the sitemap.xml to discover pages. If your sitemap was incomplete or missing, the crawler would miss content.
The docs now say you can provide "the sitemap.xml file of the source site or the URLs to specific website pages." So you can give Pathway a curated list of URLs instead of depending on the sitemap. Useful when you want to migrate specific sections of a site, or when your sitemap does not include everything.
.NET 10.0 for the Sitecore Website Path
A smaller change, but if you are following the Sitecore Website migration path - the XMComponentExtraction console app now requires .NET 10.0 instead of the .NET 9.0 I used during my testing. Make sure you have the updated runtime before setting up the extraction toolchain.
What Has NOT Changed
A few things remain the same based on what I can see in the docs:
- Two migration paths - Sitecore Website and Any Website are still the two options. The core workflow (extract → audit → map → migrate) has not changed.
- Target structure required first - You still need to set up your SitecoreAI site structure (templates, components, page designs, partial designs) before running Pathway. It maps to existing structures, it does not create new ones.
- Media library migration for Sitecore path - The old XM to XM Cloud Migration Tool is still needed for migrating media from Sitecore XM/XP sources.
- No page/partial design migration - The docs still say Pathway does not migrate page and partial designs, or XP-related items like xDB data, personalization, and email marketing content.
What I Want to Test Next
Reading the docs is one thing. Actually running a migration with these changes is another. Here is what I plan to test in my next hands-on session:
- The 200 URL limit - Can Pathway now crawl all my posts in a single run?
- Image import - Does the toggle actually pull blog images into the media library? How does the quality and organization look?
- URL list input - Can I feed Pathway a specific list of pages instead of the sitemap? How does the UI handle this?
- Multi-language - I will need a multilingual test site for this, but it is high on my list.
- Overall migration quality - Has the AI mapping improved? I am curious if the tool has become more reliable.
I will share the results with screenshots in my next post.
Still on My Wish List
A few things I hoped would change but have not, at least based on the docs:
- AI customization - Still no way to guide the AI's mapping decisions. You cannot tell it "these are product pages, not blog posts" before it runs.
- Re-run capability - If something fails mid-migration, you may still need to start over.
- JavaScript-rendered content - The crawler still works with static HTML. Single-page apps built with React, Angular, or Vue will not be fully captured.
That said, the changes so far address the things that made Pathway hard to use in practice. The 50 URL limit was the biggest one - it made the tool feel like a demo rather than something you could use on a real project. At 200 URLs with image import, the Any Website path is now a lot more practical.
If you tried Pathway earlier and ran into the same constraints I did, it is worth another look.
