Search built for your WooCommerce store.
Single-tenant. Edge-native. Custom data types. AI-agent ready. The search experience your customers expected, for a flat $99/month.
Classic Bike Spares ditched Doofinder for search that actually understands their catalogue.
They sell parts for British motorcycles — Norton, Triumph, BSA. Customers search by year + make + model and by OEM part number. Doofinder can't index that shape of data. ShopFlare can: 5,500+ products, 343 motorcycle fitments, OEM number trigram, alias-aware matching, plus an MCP server so AI assistants can answer "what oil filter fits a 1971 Norton Commando?" directly.
See the live store →Birch Creative sells quilting supplies. ShopFlare reads the product photos.
Threads, battings, rotary cutters — products where the answer is a colour or a fibre the title never mentions. ShopFlare's vision pipeline runs Workers AI over every photo and makes what it sees searchable, so "white quilting thread" or "100% wool batting" lands on the right product. Different store, different data shape, same single-tenant engine.
Try the Birch & Classic Bike Spares demos →Not a generic search box rented to 10,000 stores.
Every ShopFlare instance is shaped around the store it runs on. Custom data types, brand-specific synonyms, OEM number rules — all of it lives in your install, not a one-size-fits-all SaaS.
Single-tenant by default
Your store gets its own dedicated Cloudflare Worker, D1 database, and search index. Not a shared pool. Your data shape, your performance budget, your control.
Built for your data, not just products
Index products, posts, pages, categories, brands, and custom data types like motorcycle fitments (year + make + model → matching parts). OEM number search included.
AI-agent ready
Every ShopFlare install ships with an MCP server. Your catalogue becomes a tool that Claude, ChatGPT, and other assistants can query directly. Your customers can talk to your store.
OEM number search
Type any 3+ digit substring of a part number and the trigram index finds it. Search '0610' and land on product '061073'. Crucial for parts catalogues where customers know the number, not the name.
Sub-5ms autocomplete
FTS5 + trigram unified index over D1, with KV-backed prefix lookups. Cloudflare's edge runs it everywhere. Your customers see results before they finish typing.
Costs us $5/mo, costs Doofinder $300/mo
Cloudflare Workers are absurdly cheap at SME volumes. We charge a flat fee that's still less than half what the incumbents charge, and pocket none of it as infra savings.
How we stack up against the incumbents.
Doofinder, Algolia, Klevu, Searchanise are good products. They're built for breadth — thousands of stores on shared infrastructure. ShopFlare goes the other way.
| Feature | Doofinder | Algolia | ShopFlare |
|---|---|---|---|
| Starting price | USD $39/mo | Free → usage | AUD $99/mo flat |
| Tenancy | Shared SaaS | Shared SaaS | Single-tenant Worker |
| Custom data types (e.g. motorcycle fitments) | Not natively | DIY index | Yes, built in |
| OEM / part number trigram search | No | DIY | Yes |
| Vision-enriched index (read product photos) | No | No | Yes |
| MCP server for AI assistants | No | No | Yes |
| Hosting | Their cloud | Their cloud | Cloudflare's edge |
| Australian-built support | Spain | France/US | Newcastle, NSW |
Comparison reflects publicly advertised pricing and features as of May 2026. Prices in USD or EUR converted at indicative rates. Doofinder and others may match individual features under custom enterprise plans.
Your store, queryable by AI.
Every ShopFlare install ships with a Model Context Protocol (MCP) server. That means Claude Desktop, ChatGPT, the Cursor IDE — any MCP-aware assistant — can search your catalogue, get product details, find parts for a specific bike, locate manuals, and check stock-aware feasibility.
Plug your store into a conversational AI chat widget, an internal tool, or a customer's own assistant. The hard work of making your data agent-readable is already done.
Developer docs & MCP tool list →search(query, type?, limit?) get_product(id) → aliases, manuals, fitments find_parts_for_bike(year, make, model) nearest_fitment(year, make, fuzzy_model) check_fitment_availability(fitment_id) find_similar_products(product_id) find_manuals(query?, make?, sku?) list_models(make?, search?) list_brands(), list_categories() // + 3 more
Replace your $300/month Doofinder with a $99 ShopFlare.
Flat fee. One-off $499 setup. We do the WooCommerce integration, sync your products, install the widget, and tune it for your store. Usually live within a week.