Adding filters to a collection page and making the search bar return the right products are two of the most common requests on any Shopify store.

Shopify ships a free first-party app for exactly this: Shopify Search & Discovery. It is usually the first thing recommended. Yet as of August 31, 2026, its Shopify App Store rating is 2.7 across 464 reviews, with 90 one-star reviews — 19% of the total.

Free, built by Shopify itself, and still rated 2.7. The reason is not that the app is broken. It is that the limits are published in the Help Center, and merchants hit them after they have already built around the app.

This article lists those limits as numbers, defines the line where a paid search app becomes necessary, and shows why the three leading paid apps cannot be compared on monthly price — because they bill on three completely different units. Research date: August 31, 2026.

The short version

• Search itself works without any app. AI-backed storefront search, including predictive search and typo tolerance, is available by default to every merchant with an online store.
• Filters are what the app is for. Search & Discovery is free, has no usage cap, and is the correct starting point.
• Three limits cause most of the pain. 25 filters per store, no filters on collections over 5,000 products, and a maximum of 100 filter values shown per filter on the storefront.
• The decision to leave is about limits, not features. Counting whether 25 filter axes is enough answers the question faster than any feature matrix.
• The three paid apps do not line up on price. Searchanise bills on product count, Fast Simon on monthly sessions, Boost AI Search & Filter on online GMV. Two stores of the same size can rank these three in opposite orders.

What already works without the app

First, a separation. When people say "Shopify search is weak," they usually mix the app and the platform.

Per the Help Center, online store search runs on AI-powered search infrastructure by default and returns results across products, blog posts, and pages. Predictive search — suggestions as the customer types — and typo tolerance, where misspellings are matched to close words, are both standard.

So you do not install an app to make the search bar work.

What Search & Discovery adds sits on top:

• Filters on collection pages and search results
• Synonym groups, so "joggers" and "sweatpants" return the same results
• Product boosts, to force specific products up for specific queries
• Related and complementary products on product detail pages
• Search analytics

Two setup gotchas the Help Center documents: the app's features require Shopify Network Intelligence to be enabled in your privacy settings, and staff need the Products, Online Store Search and Navigation, and Reports permissions. A merchant who "can't edit filters" is often looking at a permissions problem, not a bug.

The published limits

This is the core of it. Every number below comes from the Shopify Help Center.

Subject	Limit
Filters per store	25 (standard and custom combined)
Collection size for filters to display	Filters hidden above 5,000 products
Search results size for filters to display	Filters hidden above 100,000 results
Filter values shown on the storefront	100 per filter
Values in a single filter group	200
Filter groups per store	1,000
Filter values shown inside the app	1,000
Unique tag values	5,000
Unique product option and attribute values	1,000

Standard filters cover Availability, Category, Price, Product type, Tags, and Vendor. Anything beyond that is a custom filter built on product options, metafields, or metaobjects.

Also worth knowing: each filter source can be used only once per store. Turn on the Price filter and Price is greyed out as a source for any other filter.

When 25 becomes the problem

Twenty-five is plenty for a single-category store. It breaks down when the catalog is broad.

Sell furniture, apparel, food, and homeware from one store and each category needs different refinement axes. Furniture: material, dimensions, weight capacity. Apparel: color, size, season. Food: ingredients, allergens, storage. Cover all of them properly and 25 disappears fast.

A July 2, 2026 review (Furniture Fair, United States, over 2 years using the app) calls the cap a bait-and-switch: Shopify markets a highly customizable platform, and merchants building a sophisticated storefront run into an arbitrary limit that forces them either to compromise the experience or pay for a third-party solution.

Shopify has not published any plan to raise the cap. Whether 25 is enough is something you can count on paper before installing anything. That is the first fork in the road.

The constraints that bite multilingual stores

If you sell across languages or currencies, the translation caveats matter as much as the filter count. From the Help Center:

• Translations are not supported for Vendor and Tags filter values
• The product tag filter is only displayed to customers shopping in the store's default language
• Vendor filter values are always based on the default language
• Filter value translations do not apply to market-specific language variants (French for Canada and French for France share one set of filter values)
• The price filter does not display for any currency other than the store's default

The practical consequence: build your filters on tags and they will break on every translated version of the store. The Help Center's own recommendation is to use metafield filters instead when translation matters. This is a design-time decision, not something you patch later.

Where the paid line sits

You have outgrown the free app if any of these is true:

1. You need more than 25 filter axes
2. You need filters on a collection with more than 5,000 products
3. You need more than 100 values visible on a single filter
4. You want the availability filter always on — the native app has no such setting, as a July 6, 2026 review (PIPI & PUPU and friends, Italy) points out, so shoppers filtering by shoe size still see out-of-stock results unless they also tick availability themselves
5. You need fine control over result ranking, or A/B testing of merchandising rules

If none of these apply, stay on the free app. A monthly fee buys you one of the items above and nothing else.

The three paid apps

Based on Shopify App Store listings as of August 31, 2026. Pricing changes, so confirm current figures before installing.

	Searchanise Search & Filter	Fast Simon AI Search & Filters	Boost AI Search & Filter
Developer	Searchanise	Fast Simon	Boost Commerce
Rating (reviews)	4.8 (1,071)	4.8 (286)	4.8 (1,500)
Admin languages	English only	English only	English only
Billing unit	Product count	Monthly sessions	Online GMV
Free plan	Yes (under 25 products, or dev stores)	Yes (100 monthly sessions)	No
Entry paid tier	$19/mo (to 1,500 products)	$39.99/mo (2,000 sessions)	From $29/mo
Middle tier	$39/mo (to 7,500 products)	$99.99/mo (10,000 sessions)	From $299/mo
Upper tier	$49/mo (Plus, to 1,500 products)	$299.99/mo (30,000 sessions)	From $699/mo
Free trial	14 days	Not stated	21 days

What the table actually says

First, the monthly figures are not comparable. Three apps, three billing units.

A 30,000-SKU catalog with modest traffic blows past Searchanise's highest listed tier (7,500 products) while sitting comfortably in a mid Fast Simon plan. A 50-SKU store doing 100,000 monthly sessions inverts that completely. Boost bills on GMV, so a store whose revenue grows faster than its catalog or its traffic pays more for identical usage.

One caveat on that comparison: each Searchanise tier carries a "Visit our website to see more pricing options" note, meaning higher tiers exist on the developer's own site that the App Store listing does not show. Large catalogs cannot be priced from the listing alone.

Second, Boost's range is an order of magnitude wider than the other two. The listing states Launch runs "up to $159/month for stores under $50K monthly GMV," Convert "up to $599/month for stores under $400K monthly GMV," and Accelerate "up to $1499/month for stores under $2.5M monthly GMV." Budgeting from the "$29" headline will not match the invoice — the same gap between quoted and billed amounts covered in Why Your Shopify App Bill Doesn't Match the Quote.

Third, session billing has a bot problem. A November 6, 2025 review (Wallhogs, United States, over 2 years using the app) describes it precisely: a large catalog means many pages, many pages mean heavy crawler traffic, and those visits count as sessions. The store averaged roughly 10,000 bot visits a month, then grew past its tier. The same reviewer notes that other filter apps generally bill on product count, and moved to one. If your catalog is large, model this before choosing a session-priced app.

Fourth, all three are English-only in the admin. The Languages field on each listing shows English and nothing else. Shopify Search & Discovery, by contrast, lists 20 languages. For teams that need a localized admin, the free app is the only option regardless of feature depth.

Fifth, none of the three carries a Built for Shopify badge. As covered in What the Built for Shopify Badge Doesn't Certify, the badge was never a quality ranking — but in this category it cannot even be used as a shortlist filter, because nobody has it.

Picking by situation

A few hundred products, under 10 filter axes

Stay on Shopify Search & Discovery. Paying does not solve a problem you have. Spend the effort on synonym groups, product boosts, and reviewing zero-result queries in the app's analytics instead.

Broad catalog, more than 25 filter axes

You need to move. Choose by store shape:

• Many products, moderate traffic → session-based Fast Simon usually costs less
• Few products, heavy traffic → product-count-based Searchanise usually costs less

Note that neither free tier helps you test this at scale: Searchanise's free plan is restricted to development stores or stores under 25 products, and Fast Simon's caps at 100 monthly sessions. Both are proof-of-concept tiers, not trial runs. Searchanise offers a 14-day trial for real evaluation.

A collection over 5,000 products

The native app shows no filters at all on that page. The Help Center's advice is to split large collections into smaller ones. If your navigation design cannot absorb that, a third-party app is the answer.

A non-English admin is required

The free app is the only option. Design within the limits, or staff the project with someone comfortable in an English admin.

Pre-install checklist

• Write out every filter axis you need and confirm it fits in 25
• Check whether any collection exceeds 5,000 products
• Check whether any single filter could exceed 100 values
• If you sell in multiple languages, build filters on metafields rather than tags
• Confirm Shopify Network Intelligence is enabled in privacy settings
• Grant staff the Products, Online Store Search and Navigation, and Reports permissions
• Measure your actual product count, monthly sessions, and monthly GMV before evaluating paid apps
• Run all three numbers against each app's billing unit and project 12 months
• Confirm the admin language matches your team
• Confirm your theme supports filter display (visible under Content > Menus)

Wrapping up

This category is decided by numbers, not by feature checklists.

1. Separate the platform from the app. Search works without an app. Filters, synonyms, boosts, and recommendations are what the app adds.
2. Test 25, 5,000, and 100 against your own catalog. Clear them and your cost stays at zero.
3. If you exceed them, choose on billing unit, not features. Product count, sessions, or GMV can swing annual cost several times over for the same store.
4. A localized admin narrows the field to one app. Every major paid competitor is English-only.
5. Use the trials, not the free tiers, to evaluate. Searchanise's free plan is limited to stores under 25 products and Fast Simon's to 100 monthly sessions. Searchanise offers 14 days and Boost 21 days of real trial.

All figures reflect the Shopify Help Center and Shopify App Store listings as of August 31, 2026. Both limits and pricing change, so verify against primary sources before you install.