Google ML Kit Barcode Scanning vs IronBarcode: C# Barcode Library Comparison

What Google ML Kit Barcode Scanning Is

Google ML Kit Barcode Scanning is a native mobile library. That is not a limitation to work around; it is the product's design. ML Kit ships in two flavors:

• Android: Distributed via Google Maven as com.google.mlkit:barcode-scanning (Kotlin/Java). Requires Google Play Services on the target device. Processes InputImage objects constructed from camera frames, file URIs, or bitmaps.
• iOS: Distributed via CocoaPods as GoogleMLKit/BarcodeScanning, with a Firebase dependency. Uses Swift or Objective-C.

There is no dotnet add package google-mlkit-barcode. There is no using Google.MLKit.BarcodeScanning;. Google does not publish a .NET binding for ML Kit.

Community-maintained Xamarin bindings have existed at various points, but they have a consistent problem: ML Kit updates its Android and iOS SDKs frequently, and binding projects maintained by individuals or small teams tend to lag behind or break entirely when the underlying API changes. As of 2026, there is no actively maintained, production-ready .NET NuGet package that wraps ML Kit's barcode scanner.

What ML Kit Does Well

Understanding why ML Kit appears in barcode comparisons is useful context. On native Android:

• On-device inference: No server round-trip. ML Kit's model runs locally using Google Play Services, which means low latency and no network dependency at scan time.
• Damage tolerance: The ML model handles damaged, partially obscured, or low-resolution barcodes better than many threshold-based decoders. This is a real differentiator for consumer-facing apps scanning real-world codes.
• Google Lens integration: ML Kit's scanner is the same stack that powers Google Lens barcode detection. That is a strong quality signal.
• Format breadth on mobile: QR Code, EAN-13, EAN-8, Code 128, Code 39, Code 93, Codabar, ITF, PDF417, Data Matrix, Aztec, and UPC-A/UPC-E are all supported natively.
• Zero configuration: A three-line Kotlin snippet can produce a working scanner without tuning thresholds or choosing a decoder strategy.

These are legitimate strengths. The issue for .NET developers is not that ML Kit is bad — it is that ML Kit is not available to them without significant work.

The .NET Development Reality

When a .NET developer needs to scan barcodes, the ML Kit path looks like this in practice:

No async/await C# API. ML Kit uses Android's Task API with addOnSuccessListener and addOnFailureListener callbacks. These do not map to .NET's Task<T> and await. Any binding layer must adapt between callback-based Android async and .NET's TPL — a non-trivial translation that community bindings handle inconsistently.

No .NET dependency injection integration. BarcodeScanning.getClient(options) is a static factory call that returns an Android BarcodeScanner object. There is no .NET interface to register, no IServiceCollection.AddBarcodeScanner(), no way to inject it into ASP.NET Core middleware or Azure Functions.

No ASP.NET Core or Azure Function support. ML Kit requires Android or iOS runtime context. It cannot run in a web API process on Linux, in an Azure Function, on Windows Server, or in a Docker container. If your use case is server-side barcode processing — a REST endpoint that accepts an image and returns barcode data, a document processing pipeline, a batch job — ML Kit is architecturally incompatible regardless of any binding availability.

Camera or frame input only — no file, no PDF. ML Kit's InputImage can be constructed from a Bitmap, a ByteBuffer, or a file URI on the Android filesystem. There is no concept of processing a PDF document, iterating pages, or handling a multi-page TIFF. Server-side document processing is outside the product's scope entirely.

Google Play Services dependency. The on-device ML model runs through Google Play Services. Devices without Google Play Services — custom Android builds, certain enterprise devices, Amazon Fire tablets — cannot use the default ML Kit configuration. The "bundled" model option (com.google.mlkit:barcode-scanning-bundled) works around this but increases APK size significantly.

Porting Android ML Kit Code to .NET

If you have an existing Android app with ML Kit barcode scanning and you are porting it to .NET MAUI or .NET 9, here is the translation. The Kotlin code uses ML Kit's callback pattern; the C# code uses IronBarcode's synchronous API.

// Android Kotlin: ML Kit
val options = BarcodeScannerOptions.Builder()
    .setBarcodeFormats(Barcode.FORMAT_QR_CODE, Barcode.FORMAT_CODE_128)
    .build()
val scanner = BarcodeScanning.getClient(options)
val inputImage = InputImage.fromFilePath(context, uri)
scanner.process(inputImage)
    .addOnSuccessListener { barcodes ->
        for (barcode in barcodes) {
            val rawValue = barcode.rawValue
            val format = barcode.format
        }
    }
    .addOnFailureListener { e -> Log.e("MLKit", e.message ?: "") }

// .NET C#: IronBarcode
// NuGet: dotnet add package IronBarcode
using IronBarCode;

var results = BarcodeReader.Read("captured-image.jpg");
foreach (var barcode in results)
{
    Console.WriteLine($"{barcode.Format}: {barcode.Value}");
}

// .NET C#: IronBarcode
// NuGet: dotnet add package IronBarcode
using IronBarCode;

var results = BarcodeReader.Read("captured-image.jpg");
foreach (var barcode in results)
{
    Console.WriteLine($"{barcode.Format}: {barcode.Value}");
}

Imports IronBarCode

Dim results = BarcodeReader.Read("captured-image.jpg")
For Each barcode In results
    Console.WriteLine($"{barcode.Format}: {barcode.Value}")
Next

The structural difference is more than syntax. The ML Kit version sets up a scanner object, constructs an InputImage, calls scanner.process(), and registers success and failure callbacks. If you need the result before proceeding, you have to coordinate callback execution with the rest of your logic — typically through a CountDownLatch in Java or coroutines in Kotlin.

IronBarcode's BarcodeReader.Read() returns a result collection synchronously. You iterate it immediately. There is no callback registration, no thread synchronization, no separate scanner object to manage.

For teams porting Android code, the pattern translation for multi-barcode scenarios looks like this:

// .NET C#: IronBarcode — reading multiple barcodes with options
using IronBarCode;

var options = new BarcodeReaderOptions
{
    Speed = ReadingSpeed.Balanced,
    ExpectMultipleBarcodes = true,
    ExpectedBarcodeTypes = BarcodeEncoding.QRCode | BarcodeEncoding.Code128
};

var results = BarcodeReader.Read("image.jpg", options);
foreach (var barcode in results)
{
    Console.WriteLine($"Format: {barcode.Format}, Value: {barcode.Value}");
}

// .NET C#: IronBarcode — reading multiple barcodes with options
using IronBarCode;

var options = new BarcodeReaderOptions
{
    Speed = ReadingSpeed.Balanced,
    ExpectMultipleBarcodes = true,
    ExpectedBarcodeTypes = BarcodeEncoding.QRCode | BarcodeEncoding.Code128
};

var results = BarcodeReader.Read("image.jpg", options);
foreach (var barcode in results)
{
    Console.WriteLine($"Format: {barcode.Format}, Value: {barcode.Value}");
}

Imports IronBarCode

Dim options As New BarcodeReaderOptions With {
    .Speed = ReadingSpeed.Balanced,
    .ExpectMultipleBarcodes = True,
    .ExpectedBarcodeTypes = BarcodeEncoding.QRCode Or BarcodeEncoding.Code128
}

Dim results = BarcodeReader.Read("image.jpg", options)
For Each barcode In results
    Console.WriteLine($"Format: {barcode.Format}, Value: {barcode.Value}")
Next

The ExpectedBarcodeTypes flag is the equivalent of ML Kit's setBarcodeFormats(). Setting it narrows the search and improves performance, but unlike ML Kit, leaving it unset does not break the read — IronBarcode will attempt all supported formats.

Format Coverage Comparison

Both tools cover the mainstream 2D and 1D formats. The overlap is substantial.

IronBarcode supports 50+ encoding types across both reading and generation. ML Kit's format list is fixed to what Google ships with the model — you cannot add custom symbologies.

What IronBarcode Adds

Beyond reading, IronBarcode covers the full barcode workflow:

Generation. ML Kit has no generation API at all — it reads barcodes, it does not create them. IronBarcode generates any supported format to PNG, JPEG, SVG, HTML, or binary data.

// Generate a Code 128 barcode to file
BarcodeWriter.CreateBarcode("SHIP-20240312-7834", BarcodeEncoding.Code128)
    .SaveAsPng("shipping-label.png");

// Generate a Code 128 barcode to file
BarcodeWriter.CreateBarcode("SHIP-20240312-7834", BarcodeEncoding.Code128)
    .SaveAsPng("shipping-label.png");

' Generate a Code 128 barcode to file
BarcodeWriter.CreateBarcode("SHIP-20240312-7834", BarcodeEncoding.Code128) _
    .SaveAsPng("shipping-label.png")

QR code customization. QRCodeWriter supports logo embedding, color changes, and error correction levels — features that have no ML Kit equivalent because ML Kit does not generate anything.

// QR code with embedded logo and custom color
QRCodeWriter.CreateQrCode("https://example.com/product/4821", 500)
    .AddBrandLogo("company-logo.png")
    .ChangeBarCodeColor(System.Drawing.Color.DarkBlue)
    .SaveAsPng("product-qr.png");

// QR code with embedded logo and custom color
QRCodeWriter.CreateQrCode("https://example.com/product/4821", 500)
    .AddBrandLogo("company-logo.png")
    .ChangeBarCodeColor(System.Drawing.Color.DarkBlue)
    .SaveAsPng("product-qr.png");

' QR code with embedded logo and custom color
QRCodeWriter.CreateQrCode("https://example.com/product/4821", 500) _
    .AddBrandLogo("company-logo.png") _
    .ChangeBarCodeColor(System.Drawing.Color.DarkBlue) _
    .SaveAsPng("product-qr.png")

PDF processing. BarcodeReader.Read("document.pdf") reads barcodes from every page of a PDF natively. No image extraction step, no page-by-page loop with a separate PDF library — it handles PDFs directly.

// Read all barcodes from every page of a PDF
var results = BarcodeReader.Read("invoice-batch.pdf");
foreach (var barcode in results)
{
    Console.WriteLine($"Page {barcode.PageNumber}: {barcode.Value}");
}

// Read all barcodes from every page of a PDF
var results = BarcodeReader.Read("invoice-batch.pdf");
foreach (var barcode in results)
{
    Console.WriteLine($"Page {barcode.PageNumber}: {barcode.Value}");
}

Imports System

' Read all barcodes from every page of a PDF
Dim results = BarcodeReader.Read("invoice-batch.pdf")
For Each barcode In results
    Console.WriteLine($"Page {barcode.PageNumber}: {barcode.Value}")
Next

Server-side and cloud. IronBarcode runs on Windows, Linux, macOS, Docker, Azure Functions, and AWS Lambda. There is no Google Play Services requirement, no camera, no mobile runtime.

Binary output. .ToPngBinaryData() returns a byte[] directly — useful for APIs that return barcode images as HTTP responses without writing to disk.

// Generate barcode and return as byte array (e.g., in a web API)
byte[] barcodeBytes = BarcodeWriter.CreateBarcode("ORDER-8821", BarcodeEncoding.QRCode)
    .ToPngBinaryData();

// Generate barcode and return as byte array (e.g., in a web API)
byte[] barcodeBytes = BarcodeWriter.CreateBarcode("ORDER-8821", BarcodeEncoding.QRCode)
    .ToPngBinaryData();

' Generate barcode and return as byte array (e.g., in a web API)
Dim barcodeBytes As Byte() = BarcodeWriter.CreateBarcode("ORDER-8821", BarcodeEncoding.QRCode) _
    .ToPngBinaryData()

Feature Comparison

API Concept Mapping

Because there is no official .NET ML Kit API, this mapping is conceptual — it shows the intent of each ML Kit operation and the IronBarcode equivalent.

When to Consider IronBarcode

Porting an Android app to .NET MAUI or .NET 9. If your Android app uses ML Kit for barcode scanning and you are building a .NET equivalent, IronBarcode is the natural target. The concept translation is straightforward — read a file, get a result, iterate values.

Server-side barcode processing. Any scenario involving an HTTP endpoint that scans barcodes, a background job processing uploaded images, or a document workflow that extracts codes from PDFs. ML Kit cannot participate in any of these.

Windows desktop applications. WinForms, WPF, or .NET MAUI on Windows. ML Kit does not run on Windows at all.

Environments without Google Play Services. If your Android deployment includes devices that do not have Google Play Services — enterprise hardware, custom AOSP builds, Amazon Fire devices — the standard ML Kit model is unavailable. IronBarcode has no such dependency.

Generation alongside reading. If you need to both generate and read barcodes — print shipping labels and verify scans, for example — IronBarcode handles both in the same library. ML Kit handles neither generation nor server-side reading.

Azure or AWS deployment. Azure Functions, Azure App Service on Linux, AWS Lambda — none of these can host ML Kit. IronBarcode targets all of them.

Licensing and Pricing

Google ML Kit Barcode Scanning is free for Android and iOS apps. There is no per-scan fee and no commercial license required. That pricing model makes sense in its intended context — a native mobile library bundled with Google Play Services.

IronBarcode is a commercial .NET library with perpetual licenses:

• Lite: $749 — 1 developer, 1 project location
• Plus: $1,499 — up to 3 developers
• Professional: $2,999 — up to 10 developers
• Unlimited: $5,999 — unlimited developers

A free trial is available without a time limit; the trial watermarks generated barcodes. For teams already paying for server infrastructure to process documents, the license cost is typically a small fraction of the engineering effort required to maintain an unofficial ML Kit binding.

Conclusion

Google ML Kit Barcode Scanning is a well-engineered library that does exactly what it was designed to do: scan barcodes quickly and accurately on Android and iOS devices. The comparison with IronBarcode is less about which library is better and more about which library exists in the context where you need it.

If you are writing Kotlin for Android and scanning from a camera, ML Kit is an excellent choice. If you are writing C# for .NET — whether for a web API, a document processing service, a Windows application, a cloud function, or a cross-platform MAUI app — Google ML Kit has no offering for you. There is no NuGet package, no C# API, and no Google-supported path to using it in managed code. IronBarcode is a native .NET library that covers both reading and generation across every platform .NET supports.