GitHub Copilot vs. Amazon CodeWhisperer: What Developers Need to Know

On the surface, GitHub Copilot and Amazon CodeWhisperer don't look that different. They are both tools that use AI to help accelerate software development processes. They are also both owned by and integrated with the ecosystems of major tech companies — Microsoft in the case of Copilot, and Amazon in the case of CodeWhisperer — which makes them different from AI-assisted development tools owned by startups, such as Tabnine, rather than tech behemoths.

But when you dive into the details, you'll notice that Copilot and CodeWhisperer work a bit differently, as this article explains. They also focus on different sets of use cases, which is the most important distinction between them.

What Is GitHub Copilot?

Copilot is an AI-assisted software development tool from GitHub (which is owned by Microsoft). GitHub introduced it in 2021 and released it into general availability in 2022.

Copilot works by analyzing the code that developers have written — including comments that describe how code should work, even if the code itself isn't written yet — then offers automated suggestions for writing new code. Under the hood, Copilot is powered by OpenAI Codex, an AI model that was trained on code inside millions of publicly available source code repositories.

Related:4 Ways AI-Assisted Coding Can Benefit ITOps Engineers

What Is Amazon CodeWhisperer?

CodeWhisperer is an AI-assisted software development tool from Amazon. Like Copilot, CodeWhisperer interprets comments made by developers, then automatically suggests code they might want to implement.

Amazon hasn't disclosed a lot of details about how CodeWhisperer works in the back end, but it says that the AI models that power the tool were trained using both open source code and code that is internal to Amazon.

GitHub Copilot vs. Amazon CodeWhisperer: Key Similarities

From a functional perspective, Copilot and CodeWhisperer are quite similar. They both address the same needs, and developers can use them in the same ways. With either tool, you basically just describe what you want your code to do, then let AI models auto-suggest the code you need to do it.

Copilot and CodeWhisperer are also similar in that they are both cloud-based solutions offered by major tech vendors.

A third similarity is that neither tool claims to be capable of writing entire software programs on its own — at least for now. Both Copilot and CodeWhisperer limit themselves to helping complete relatively small segments of code based on context that developers provide.

Major Differences Between Copilot and CodeWhisperer

When you dig deeper into the specific use cases for each tool, however, you'll notice some important differences between GitHub Copilot and Amazon CodeWhisperer.

Related:Does AI-Assisted Coding Violate Open Source Licenses?

The biggest difference is that Copilot is designed to be more of a general-purpose AI-assisted development tool, whereas CodeWhisperer caters first and foremost to development use cases associated with Amazon platforms, such as Amazon Web Services.

To be clear, this doesn't mean you can't use CodeWhisperer to help write applications that aren't linked in any particular way to the Amazon ecosystem. You certainly can. Nor does it mean that code written with CodeWhisperer will only run on Amazon platforms. In most cases, it can run anywhere.

Still, when it comes to writing code related to Amazon technologies, CodeWhisperer typically does a better job than Copilot. If you want to write code for moving files between Amazon S3 buckets, for instance, or for working with EC2 instances, you'll probably have an easier time getting good code if you use CodeWhisperer.

In contrast, although Copilot is hosted on a Microsoft-owned platform, it doesn't cater in any special way to Microsoft technologies or Microsoft-related programming use cases. It's a general-purpose tool.

Another important difference between Copilot and CodeWhisperer is that CodeWhisperer supports many fewer programming languages and IDEs. Currently, it's compatible with just C#, Java, JavaScript, Python, and TypeScript as programming languages, and most of the IDEs it supports are Amazon-based ones (JetBrains and Visual Studio Code are the exceptions).

Copilot isn't subject to these limitations. Copilot can generate code in virtually any language. And although GitHub says it's "optimized" only for certain languages, the list of such languages — which include Python, JavaScript, TypeScript, Ruby, Go, C#, and C++ — is longer than the list of languages that CodeWhisperer supports. Copilot also supports almost all of the major IDEs.

Choosing Between GitHub Copilot and Amazon CodeWhisperer

Because of the differences described above, it seems relatively easy at present to decide whether Copilot or CodeWhisperer is better for varying AI-assisted development needs.

In general, Copilot is the clear choice because it offers support for a much broader set of use cases. At present, the main reason to choose CodeWhisperer over Copilot would be if you are writing code related to Amazon APIs, and/or if you only need your AI-assisted development tool to support the limited set of programming languages and IDEs with which CodeWhisperer is compatible.

This may change in the future if Amazon extends CodeWhisperer into a more general-purpose AI-assisted development platform. But for now, Copilot beats CodeWhisperer when it comes to flexibility and breadth, although CodeWhisperer is better at writing code for Amazon's own APIs.