Streaming Tasks with .NET 9 Task.WhenEach

Handling multiple parallel tasks has always been a bit binary in .NET. You either wait for everything to finish (Task.WhenAll) or you wait for the first one (Task.WhenAny).

But what if you want to process results as they arrive?

Before .NET 9, you had to write a complex loop involving Task.WhenAny, removing finished tasks from a list, and looping again. This was not only verbose but had O(N^2) performance characteristics.

Enter Task.WhenEach

In .NET 9, Task.WhenEach solves this natively. It returns an IAsyncEnumerable that yields tasks as they complete.

using System.Threading.Tasks;

public async Task ProcessImagesAsync(List<string> urls)
{
    // Start all downloads in parallel
    List<Task<byte[]>> downloadTasks = urls.Select(DownloadAsync).ToList();

    // Process them one by one, as soon as they finish
    await foreach (var completedTask in Task.WhenEach(downloadTasks))
    {
        try 
        {
            byte[] image = await completedTask;
            Console.WriteLine($"Processed image: {image.Length} bytes");
        }
        catch (Exception ex)
        {
            Console.WriteLine($"One download failed: {ex.Message}");
        }
    }
}

Why This Matters

1. User Experience: In UI apps (MAUI, WPF), you can render items progressively instead of showing a spinner until the entire batch is done.
2. Throughput: In backend services, you can start processing the first results immediately, keeping the CPU fed while waiting for slower I/O operations to finish.
3. Simplicity: The code is linear and readable. No more while loops or list mutations.

This is one of those small API additions that deletes a lot of “utility” code from our projects.