Is there any difference between using AVPlayer, AVPlayerItem, and AVPlayerAsset?

I found the three ways to play video with URL.

let url = "some url"

// first way
AVPlayer(url: url)

// second way
let playerItem = AVPlayerItem(url: url)
AVPlayer(playerItem: playerItem)

// third way
let asset = AVAsset(url: url)
let playerItem = AVPlayerItem(asset: asset)
AVPlayer(playerItem: playerItem)

Is there any difference between above these?

From the docs of AVPlayer.init(url:)

This method implicitly creates an AVPlayerItem object. You can get the player item using currentItem.

So we know that when you use the first way, something similar to the second way is happening under the hood - an AVPlayerItem will be created with the URL you specified. Therefore, the first and second way are the same.

Although the docs doesn't say this explicitly, I'm pretty sure AVPlayerItem.init(url:) creates an AVAsset using the URL you specified too, since an AVPlayerItem is:

An object used to model the timing and presentation state of an asset played by the player.

The wording suggests that you can't have a AVPlayerItem without a AVAsset. Indeed, AVPlayerItem.asset is a non-optional property. So you really need an AVAsset to create an AVPlayerItem. That, combined with the fact that AVPlayerItem.init(url:) is a convenience initialiser, and init(asset:automaticallyLoadedAssetKeys:) is the designated one, I am quite sure AVPlayerItem.init(url:) creates an AVAsset under the hood as well.

If you are wondering why AVPlayer.init(playerItem:) and AVPlayerItem.init(asset:) exist when all three ways do the same thing anyway, they are for when you just so happen to be working with AVPlayerItems when you want to create a AVPlayer, and when you just so happen to be working with AVAssets when you want to create an AVPlayerItem.