Medusa.js does not ship a point-of-sale screen. It provides the commerce backend a POS needs (products, inventory, stock locations, sales channels, customers, promotions) through APIs, and the screen staff use at the counter is a separate app built on top. That is the first thing to understand before you evaluate "Medusa POS" for a store.
Retailers searching for "Medusa POS" often expect a product they can install and open, like Shopify POS or a till system. Medusa is closer to an architecture: a headless commerce backend with the modules a POS needs, and on top either an open-source starter app or a front end built for your store.
This is a different model, with real trade-offs on both sides. I lead ERPNext and Frappe work at MithTech, a team of Frappe specialists in Bengaluru who build mainly on Frappe and Medusa. Our own counter app on a Medusa backend is CounterFlow, an offline-first POS.
Does Medusa.js come with a built-in POS?
No. Medusa provides the backend (inventory, stock locations, sales channels, customers, promotions) through its APIs, and a separate front end is what staff use at checkout. In an all-in-one POS product the till software and the backend are one purchase; in Medusa they are not.
So "getting Medusa POS" means one of two things. You start from an open-source starter, such as Agilo's Medusa POS Starter, or you build a POS front end for your own counter workflow. Medusa's POS recipe describes the second path, including barcode lookup through the variant's barcode field and a custom API route. Either way you need a POS application layer.
Skimmable summary: Medusa provides the commerce APIs a POS needs, not the checkout screen; that comes from an open-source starter or an app built on Medusa's backend.
What does Medusa provide for a POS?
Four building blocks. Inventory and stock locations track stock across warehouses and stores. Sales channels separate online from in-store sales against the same catalogue. Customers let the cashier find a buyer and place the order on their account. Promotions apply discounts during an in-store sale.
The effect: the customer's online orders, their account and the stock figure are the same data the POS reads. No sync job reconciles an online store against a separate till; both query one backend. That is the strength of the model for a brand selling through more than one channel.
Skimmable summary: Medusa's inventory, stock location, sales channel and customer modules give a POS app the same data as your online store; the value is one stock and customer record, not a ready checkout screen.
Medusa POS or ERPNext's own POS?
Choose a POS on Medusa if you sell through more than one channel alongside your shops (an online store, an app, a B2B portal, marketplaces) or plan to. Choose ERPNext's POS if the business is retail-only and you want billing, stock and accounts inside one ERP with no separate commerce backend to run.
Neither is better in general; they solve different problems. A single shop with no online plans is usually over-building with Medusa. A brand running a D2C site, a wholesale portal and stores from shared stock is under-served by a POS that only knows the ERP.
Neither works offline by default. ERPNext's POS needs its server (details); a Medusa POS app works offline only if the app is built to store sales locally and sync later.
| Better fit | |
|---|---|
| Retail-only, no online store | ERPNext's POS |
| Online store and shops sharing stock | A POS on Medusa |
| B2B portal, retail and online from one catalogue | A POS on Medusa |
| Everything inside one ERP, no separate commerce layer | ERPNext's POS |
| Shops with unreliable internet | Either, with an outage plan: a fallback line, or a POS app built to work offline |
Skimmable summary: ERPNext's POS suits retail-only businesses that want one ERP; a POS on Medusa suits brands whose online store, B2B and shops share stock. The deciding factor is how many channels need the same stock and customer data.
How does a Medusa POS compare to Shopify POS or Lightspeed?
A POS on Medusa is open source and self-hosted, so you own the data and the code. Shopify POS and Lightspeed are hosted services: quicker to start, with subscription and payment fees set by the vendor (check their current pricing pages). If you want the data on your own infrastructure and one backend for every channel, Medusa fits; if you need a working till next week and little else, a hosted POS is the shorter path.
What hardware does a store need?
A tablet or small PC, a receipt printer, a barcode scanner and a cash drawer that opens on print. A browser-based POS app runs on iPad, Android or Windows; check which printers and scanners the app you choose supports. For several counters, plan a printer per counter and a small UPS so a power blip doesn't stop a sale halfway.
Frequently asked questions
+Does Medusa.js have its own POS app I can install?
Not one from Medusa itself. Open-source starters exist, including Agilo's Medusa POS Starter, with barcode scanning, checkout, cart, customer profiles and mobile support. Using a starter, or building a POS front end on Medusa's APIs, is how a Medusa POS is delivered.
+What's the difference between a Medusa POS and ERPNext's POS?
ERPNext ships its own POS inside the ERP, for retailers who want billing, stock and books in one system; it needs a connection to its server. A Medusa POS is an app on Medusa's commerce APIs, and its advantage is one stock and customer record across online, B2B and in-store channels. They solve different problems.
+Is a Medusa POS a good fit for a single store with no online presence?
Usually not the simplest choice. A single store with no online plans is better served by a retail POS product or ERPNext's POS. Medusa's multi-channel design solves a problem that store doesn't have.
+Can a Medusa POS work offline if the store's internet goes down?
That depends on the POS app, not on Medusa. An app built to work offline queues sales on the device and syncs when the connection returns; many don't. Ask whoever builds or supplies your front end, and test it with the line unplugged before go-live.
Comments & ratings
No comments yet. Start a new discussion.
Manoj · Published on 17 July 2026
Worked with us? Leave a Google review