Shopify's filter limits: the actual numbers, what hits them first, and your options

If your store's filters have started misbehaving — a brand filter that's missing half your brands, a big collection that suddenly shows no filters at all — you haven't broken anything. You've hit a documented ceiling. Shopify's native filtering (the Search & Discovery app) has hard limits, they're scattered across help pages, and most merchants meet them one surprise at a time.

Here they all are in one place, what tends to break first, and an honest read on your options — including the ones that cost nothing.

The documented limits (as of September 2026)

All of these come from Shopify's own documentation. Shopify changes limits over time, so treat the date on this article as part of the data:

• 25 filters per store, standard and custom combined — shared across every collection and search page. There is no per-collection filter set.
• 100 filter values displayed per filter on the storefront. If your Color filter has 140 colors, shoppers see 100 of them; the rest silently don't exist.
• Filter groups: 200 unique values per group, and 1,000 filter groups per store across all filter settings.
• Collections with more than 5,000 products show no filters at all. Not fewer filters — none.
• Searches returning more than 100,000 results show no filters.
• Tag-based filtering handles up to 5,000 unique tag values; product options and attributes, up to 1,000 unique values.

None of these are bugs, and no theme setting changes them. A theme decides how filters look; Shopify decides what filtering exists. (Our own theme, QuillCommerce, renders whatever filters the platform provides — these ceilings apply to every theme equally.)

What breaks first on a large catalog

In practice the limits bite in a predictable order:

The 100-value display cap goes first. Any store with a serious brand or vendor list crosses 100 values without noticing. The filter still renders, so nothing looks broken — shoppers just can't reach the brands that didn't make the cut. This is the most dangerous limit because it fails silently.

Then the 25-filter budget runs out. Because filters are store-wide, every department competes for the same 25 slots. A store selling rods, apparel and electronics can't give each category its own sensible facets — length and action for one, size and fit for another — without rationing.

Then tag sprawl catches up. Years of ad-hoc tagging ("sale", "SALE", "sale-2024", "clearance") burn through unique-value budgets and turn filters into noise. Tags were never designed to be a product schema; on large catalogs they get treated like one.

And the 5,000-product cutoff ends the conversation. An "All products" or top-level category collection on a large store simply loses filtering entirely — usually the exact page where shoppers need it most.

The free fixes: raise the ceiling before you buy anything

Before any app or custom build, three cleanups reclaim real headroom:

1. Consolidate tags ruthlessly. Merge duplicates and casing variants, delete dead campaign tags, and stop encoding structured data ("material: cork") in freeform tags.
2. Move real attributes into options and metafields. Filters built on structured data are precise, translatable and don't sprawl. This is also the work that improves your data for every future system, whatever you choose.
3. Split mega-collections. Anything approaching 5,000 products needs subdivision anyway — for shoppers, not just for the filter engine.

If your catalog is a few thousand well-structured products, this may be all you ever need. We tell merchants that plainly, because the next steps cost money.

When cleanup isn't enough: what a search app actually changes

A real search-and-filter app doesn't raise Shopify's limits — it stops using Shopify's filter engine altogether. The products are synchronized into the app's own search index, and search, facets and counts are computed there.

We're building one — FindSmith — so this is first-hand engineering, not app-store marketing: each store gets its own isolated index (one merchant's synonyms and boosts never leak into another's results), facets run on vendor, type, tags, variant options, availability and price without the 25-slot budget, and search cost stops scaling with catalog size. On our test bench, searches across 20,000 products return in about 6ms at p95, facet counts included — against Shopify-scale catalogs where an in-process approach measured 305ms on identical data. Typo tolerance and per-store synonyms come with the same index.

The honest trade-offs: an app is a dependency and a subscription; sync has to be engineered well (webhooks, bulk operations, reconciliation) or you get stale results; and if the app's index goes down, you want a design that degrades loudly rather than silently — not one that quietly falls back to hammering your database.

When it's custom-software territory

If your selling model itself doesn't fit faceted search — dealer-specific catalogs where each account sees different products and prices, quick-order flows over variant tables, configurators — no filter app fixes that, because the problem isn't search. That's e-commerce engineering, and it starts from your workflow, not from a plugin.

The decision line

• Filters missing values or gone entirely → check the limits above; you've probably hit one.
• Catalog under a few thousand SKUs, messy tags → cleanup first. Free, and improves everything downstream.
• Large catalog, real brand/attribute depth, shoppers who search → a dedicated search index, whether ours or another good one.
• The catalog isn't the problem, the selling model is → custom engineering.

Not sure which side of the line you're on? Describe your catalog and what shoppers can't find — an engineer reads it and gives you a straight answer within one business day, including when the answer is "just clean up your tags."