ImageEngine vs imgix — A Pricing and Performance Comparison
This is part two of a comparison between two similar services for optimizing images; ImageEngine and imgix. You’ll find part I here.
Both ImageEngine and imgix are pure image CDN platforms mainly aimed at improving website performance by optimizing images and image content and accelerating delivery.
Images make up more than half of website payloads today and are one of the main causes of slow loading pages, according to Google.
Both CDNs claim to do most of the heavy-lifting themselves, drastically reducing payloads by automatically formatting, compressing, and resizing images for optimal performance. A global CDN with multiple PoPs is also supposed to lead to higher cache-hit ratios and speed up delivery around the world.
These are the claims we are about to test. We’ll also draw a quick comparison of pricing to see how much bang you get for your buck.
Performance
In terms of performance, both image CDNs will do everything within their capabilities to automatically optimize your image content and accelerate delivery without your intervention. In Part I of this comparison, we looked at details regarding what each does to achieve this.
Now, we’ll check out the results you can expect using either ImageEngine or imgix and compare it to a standard web page not using a CDN. Considering the trend towards mobile-first indexing and the generally more challenging conditions of optimizing images for mobile, we are mainly comparing mobile results.
However, to level the playing field somewhat, the CDN-less page does use responsive image syntax.
First, let’s look at how the unoptimized page does:
While a score of 73 isn’t bad, there is still a lot of room for improvement, especially when it comes to our largest image(s) and overall loading speeds. If we look at the opportunities for improvement, we can also see that not using appropriate image formats and sizes are still an issue (even using responsive image syntax):
So, now let’s see how ImageEngine performs:
As you can see, the overall speed index improved, at least partly thanks to the largest image content also loading in a shorter time. As a result, the overall score is now a blazing 96. As you can see from the picture below, ImageEngine addressed both the image-related issues by using next-gen formats and properly sizing images:
Lastly, let’s see what imgix does under the same conditions:
Running a mobile test, imgix barely manages any improvement over simply using responsive syntax. Thanks to delivery acceleration via a CDN, it still improved the LCP. However, the reason for the low score seems to be that imgix didn’t use the optimal image formats and sizes:
This result is somewhat surprising since, in the desktop audit, imgix did address these opportunities for improvement and scored similarly to ImageEngine.
For example, the original payload for the entire page was 2,877 kB. ImageEngine reduced this to 800 kB for mobile (72% payload reduction) and 690 kB for desktop (76% payload reduction). On the other hand, imgix didn’t make a dent in the mobile payload reduction while it reduced the desktop payload to 1240 kB (57% payload reduction).
Pricing Comparison
In terms of pricing, these two Image CDN services make use of two very different models. ImageEngine uses a two-tiered pricing schedule with preset usage limits. On the other hand, imgix uses a pay-as-you-go system whereby you’ll be billed separately each month based on your usage for that period.
The obvious tradeoff between these two approaches is the ability to budget your Image CDN spend with ImageEngine vs the flexibility to only pay as much as you use with imgix.
Let’s start off with ImageEngine. ImageEngine bases its usage metrics on what they call “Smart Bytes.” This means they’ll measure your usage according to the amount of optimized image data delivered via their CDN.
For example, let’s say the image content on your website would usually result in 1 TB of data transfer per month. If we go with ImageEngine’s claim to optimize image content by up to 80%, that means your usage will only accrue to roughly 200 GB if ImageEngine’s firing on all cylinders.
With that in mind, the Basic plan will set you back $49 a month and comes with 100 GB of Smart Bytes, or roughly 400–500 GB of unoptimized image data. The Standard plan comes with 250 GB of Smart Bytes or roughly 1 TB to 1.25 TB of unoptimized image data.
If you do happen to go over these limits, overage charges apply of $0.4 for every extra GB of Smart Bytes used.
Imgix’s pay-as-you-go pricing is fairly straightforward. You pay $3 for every 1,000 master images as well as another 8ȼ per GB of CDN bandwidth. Here’s what it means:
- Master images: These are the original, unoptimized images that you want to optimize and serve via the imgix platform. Multiple optimized copies are generated for each image, but you’ll only be charged per original.
- CDN bandwidth: This is the total data transfer of optimized derivative image content served via the imgix CDN, similar to Smart Bytes but without the fancy name.
However, imgix has a minimum usage limit for each month of $10, inclusive of your master image count and bandwidth usage. If you use less than this, you’ll still be charged $10 to keep your account active.
Drawing a direct comparison between the two is tricky, especially as imgix’s pay-as-you-go pricing means it’s highly dependent on your exact situation. The following two tables will illustrate why this is. Each table is weighted according to the highlighted row:
To keep the math manageable, we based these comparisons on image-heavy sites with relatively high-quality images. For high-traffic, image-heavy, high-res image websites, ImageEngine seems to be the more affordable option.
However, the picture looks dramatically different as you reduce the number of images and image sizes. At smaller scales, imgix’s pay-as-you-go pricing starts to overtake ImageEngine in terms of bang for your buck and overall affordability.
That being said, at extremely high overages, ImageEngine’s $0.4 per GB of overage charges might start to catch up with you. However, you’d have to use something ludicrous like 1000x of the outbound data transfer than the limits allow.
The bottom line is that it’s simply impossible to say with certainty how many master images any customer would use, never mind their average size. In the end, it may simply come down to whether you prefer an all-inclusive or pay-as-you-use pricing model.
To get started, both companies offer a free trial period. With ImageEngine, you get 60 days to use the platform for free while imgix provides you with $10 of free credit that you can use until it runs out. Both should give you ample time and freedom to establish whether you want to go forward.
Conclusion
In terms of performance, there’s little doubt leftover that ImageEngine is the top choice to optimize image content. It did live up to its claims of providing a nearly 80% payload reduction on both desktop and mobile devices. When it comes to mobile, ImageEngine’s advanced device-detection capabilities paid off and imgix was no match.
That being said, you can use imgix’s extensive API to manually optimize images for mobile display further. However, you will sacrifice one of the main advantages of using an image CDN — scalability of implementation and ability to save you time in development.
When it comes to pricing, things aren’t nearly as clearcut. imgix’s pay-as-you-use model is light years different from ImageEngine’s predefined packages. imgix’s flexible pricing is attractive and it may be the more affordable option on smaller scales. However, ImageEngine’s predefined prices that are completely independent of your master image count will provide more stability and should be more affordable for sites with huge numbers of images.