Cattura vs Zebra OCR Wedge: technical differences and when to use each
If you run Zebra terminals and need to read text beyond barcodes, you have two native paths for the hardware you already own: OCR Wedge, the OCR feature included in DataWedge (Mobility DNA), and Cattura, the data-capture app built on the Zebra AI DataCapture SDK. Both are serious solutions — one is Zebra's own, the other is Zebra Validated — but they solve different problems. This page lays out the real technical differences, without marketing, to help you choose based on your use case.
Where they come from: two different architectures
OCR Wedge is a feature of the Workflow plugin in DataWedge, the Mobility DNA tool preinstalled on every Zebra terminal. You enable it by configuring a DataWedge profile: it is a layer integrated into the system wedge, not a separate app. Its strength: it is already on board and configured with the tools a Zebra administrator already knows.
Cattura is an app built on the Zebra AI DataCapture SDK, Zebra's new-generation AI engine that runs vision models directly on the terminal's Hexagon DSP/NPU (Qualcomm). For this reason Cattura requires terminals with the QCM6490, QCM5430 or QCM6690 chipset — the recent enterprise range: TC53/TC58, TC73/TC78, TC501/TC701, ET60/ET65, ET401, EM45 — and does not run on the "e" models (QCM4490) or on legacy devices without a DSP.
In short: OCR Wedge is the path integrated into DataWedge; Cattura is the path that leverages Zebra's latest on-device AI engine. The choice depends on what you need to read and how.
Difference 1 — template-based OCR vs free-text OCR
This is the most important difference in practice.
OCR Wedge works with predefined use-case models: a fixed set of data types that Zebra has trained recognition for — VIN (vehicle chassis numbers), license plates, tire identification numbers (TIN), meter readings, container codes, identity documents and driver's licenses. If your use case is one of these, OCR Wedge covers it well, out of the box.
Cattura performs free-text OCR on printed text: lots, serial numbers, expiry dates, internal codes, any readable string on a label or a component. Data quality is governed with configurable filters — regex, length, character whitelist — that discard spurious reads before they reach your back-end system.
The question to ask is simple: does the data I need to read fall within the OCR Wedge templates? If yes (e.g. a vehicle yard reading VINs), OCR Wedge is the natural answer. If instead you need to read the lot printed by your supplier, an expiry date in a non-standard format or an engraved serial — that is, text no generic template knows — you need free-text OCR, and that is Cattura's ground.
Difference 2 — barcode and OCR: separate or in a single gesture
With OCR Wedge, OCR is a flow of its own: barcode reading stays with the DataWedge barcode engine, configured separately. If the operator has to capture a barcode and some text from the same package, those are two operations with two configurations.
Cattura unifies 1D/2D barcode and OCR in the same operation: one frame, one gesture, both engines active on the same preview. For flows where the package carries barcode and text together — the lot printed above the GS1, the expiry next to the EAN — the saving is one step per item. Across thousands of items a day, it is the difference you see at the end of the shift.
Difference 3 — output to applications
Both inject the read value into the active field via keyboard wedge: they work with any app — ERP, WMS, browser, terminal server — with no changes.
Cattura adds two things: a native Intent broadcast with structured fields (value, symbology, confidence, timestamp) for Android apps that want a clean integration via BroadcastReceiver, and the floating button always visible above any app — the operator starts a scan without switching context, from any screen.
Difference 4 — licensing model
OCR Wedge is licensed per individual OCR feature, term-based (typically 1–2 years), for each data type: the VIN license is one, the license-plate license another.
Cattura has a single annual per-device license that covers everything: barcode, free-text OCR, multi-code batch, intent, updates — €135/device/year, public pricing, with a 7-day auto-activated demo.
If you need a single template for a few devices, the two options can be equivalent; if you need several data types or want a predictable cost for the whole fleet, the single-license model is simpler to manage and to budget.
Comparison table
| Zebra OCR Wedge | Cattura | |
|---|---|---|
| Architecture | Feature of the DataWedge Workflow plugin | App on the Zebra AI DataCapture SDK (Hexagon DSP) |
| OCR type | Predefined templates (VIN, plates, TIN, meters, containers, ID documents) | Generic printed text + regex/length/whitelist filters |
| Barcode + OCR | Separate flows (barcode via the DataWedge engine) | Unified in a single operation |
| Output | Keyboard wedge | Keyboard wedge + structured Intent broadcast |
| Scan trigger | DataWedge profile configuration | Floating button above any app |
| Required terminals | DataWedge / Mobility DNA range | QCM6490/5430/6690 chipset (TC5x/7x, TC501/701, ET6x, ET401, EM45) |
| License | Per individual OCR feature, 1–2 years per type | Single annual per device, barcode+OCR included, €135/year |
| Best for | Use cases that match the Zebra templates | Lots, serials, expiry dates, non-standard text; combined barcode+text flows |
When to choose which
Choose OCR Wedge if the data to read is exactly one of the Zebra templates (VIN, plate, container, document), the OCR flow separate from the barcode is enough for you, and you prefer to stay entirely within the DataWedge configuration you already administer.
Choose Cattura if you need to read free text (lots, serials, expiry dates, internal codes), you need to capture barcode and text in a single gesture, you want the output also as a structured intent, or you want a single license and public pricing for the whole fleet. Requirement: terminals with a Hexagon DSP — check your model.
And they are not mutually exclusive: on the same fleet you can use DataWedge for basic barcode scanning and Cattura where free-text OCR or the combined flow is needed.
Frequently asked questions
Does Cattura replace DataWedge?
No. DataWedge remains the system tool for traditional barcode scanning; Cattura complements it where you need free-text OCR, combined barcode+text capture or structured intent output. They can coexist on the same fleet.
Is Cattura a solution "against" Zebra?
No: Cattura is built on the Zebra AI DataCapture SDK and is Zebra Validated. It uses Zebra's official AI engine — the difference is at the application level: free text instead of templates, a unified flow, the licensing model.
Does OCR Wedge work on the terminals where Cattura runs?
Yes, they are recent Zebra terminals with Mobility DNA. The reverse is not true: Cattura requires the Hexagon DSP (QCM6490/5430/6690), so it does not run on "e" models (TC53e/TC58e, MC9400/9450) or legacy devices, where DataWedge is available instead.
Can I try Cattura before deciding?
Yes: every terminal starts with a full 7-day demo, activated automatically on first launch, with no registration.
Cattura turns compatible Zebra terminals into AI scanners: barcode and OCR in a single gesture, into the active field of any app. Try free for 7 days or see the supported terminals.