After I reviewed Amazon CodeWhisperer, Google Bard, and GitHub Copilot in June of 2023, CodeWhisperer might generate code in an IDE and did safety evaluations, nevertheless it lacked a chat window and code explanations. The present model of CodeWhisperer is now known as Amazon Q Developer, and it does have a chat window that may clarify code, and several other different options that could be related to you, particularly if you happen to do lots of growth utilizing AWS.

Amazon Q Developer at the moment runs in Visible Studio Code, Visible Studio, JetBrains IDEs, the Amazon Console, and the macOS command line. Q Developer additionally provides asynchronous brokers, programming language translations, and Java code transformations/upgrades. Along with producing, finishing, and discussing code, Q Developer can write unit checks, optimize code, scan for vulnerabilities, and counsel remediations. It helps conversations in English, and code within the Python, Java, JavaScript, TypeScript, C#, Go, Rust, PHP, Ruby, Kotlin, C, C++, shell scripting, SQL, and Scala programming languages.

You may chat with Amazon Q Developer about AWS capabilities, and ask it to overview your sources, analyze your invoice, or architect options. It is aware of about AWS well-architected patterns, documentation, and resolution implementation.

In keeping with Amazon, Amazon Q Developer is “powered by Amazon Bedrock” and educated on “high-quality AWS content material.” Since Bedrock helps many basis fashions, it’s not clear from the online assertion which one was used for Amazon Q Developer. I requested, and received this reply from an AWS spokesperson: “Amazon Q makes use of a number of fashions to execute its duties and makes use of logic to route duties to the mannequin that’s the greatest match for the job.”

Amazon Q Developer has a reference tracker that detects whether or not a code suggestion could be just like publicly out there code. The reference tracker can label these with a repository URL and undertaking license info, or optionally filter them out.

Amazon Q Developer immediately competes with GitHub Copilot, JetBrains AI, and Tabnine, and not directly competes with plenty of giant language fashions (LLMs) and small language fashions (SLMs) that find out about code, comparable to Code Llama, StarCoder, Bard, OpenAI Codex, and Mistral Codestral. GitHub Copilot can converse in dozens of pure languages, versus Amazon Q Developer’s one, and helps plenty of extensions from programming, cloud, and database distributors, versus Amazon Q Developer’s AWS-only ties.

Putting in Amazon Q Developer

Given the a number of environments during which Amazon Q Developer can run, it’s not a shock that there are a number of installers. The one difficult bit is signing and authentication.

Putting in Q Developer in Visible Studio Code

You may set up Amazon Q Developer from the Visible Studio Code Market, or from the Extensions sidebar in Visible Studio Code. You will get to that sidebar from the Extensions icon on the far left, by urgent Shift-Command-X, or by selecting Extensions: Set up Extensions from the command palette. Kind “Amazon Q” to seek out it. When you’ve put in the extension, you’ll have to authenticate to AWS as mentioned beneath.

Putting in Q Developer in JetBrains IDEs

Like Visible Studio Code, JetBrains has a market for IDE plugins, the place Amazon Q Developer is on the market. You’ll have to reboot the IDE after downloading and putting in the plugin. Then you definitely’ll have to authenticate to AWS as mentioned beneath. Notice that the Amazon Q Developer plugin disables native inline JetBrains full-line code completion.

Putting in Q Developer within the AWS Toolkit for Visible Studio

For Visible Studio, Amazon Q Developer is a part of the AWS Toolkit, which you’ll find it within the Visible Studio Market. Once more, when you’ve put in the toolkit you’ll have to authenticate to AWS as mentioned beneath.

Signing and authenticating Amazon Q Developer

The authentication course of is complicated as a result of there are a number of choices and several other steps that bounce between your IDE and internet browser. You used to should repeat this course of often, however the product supervisor assures me that re-authentication ought to now solely be vital each three months.

Putting in Q Developer for command line

Amazon Q Developer for the command line is at the moment for macOS solely, though a Linux model is on the roadmap and documented as a distant goal. The macOS set up is principally a obtain of a DMG file, adopted by operating the disk picture, dragging the Q file to the purposes listing, and operating that Q app to put in the CLI q program and a menu bar icon that may convey up settings and the online person information. You’ll additionally have to authenticate to AWS, which is able to log you in.

Amazon Q Developer within the AWS Console

If you’re operating as an IAM person slightly than a root person, you’ll have so as to add IAM permissions to make use of Amazon Q Developer. After getting permission, AWS ought to show an icon on the proper of the display that brings up the Amazon Q Developer interface.

Evaluating Amazon Q Developer

In keeping with AWS, “Amazon Q Developer Agent achieved the very best scores of 13.4% on the SWE-Bench Leaderboard and 20.5% on the SWE-Bench Leaderboard (Lite), an information set that benchmarks coding capabilities. Amazon Q safety scanning capabilities outperform all publicly benchmarkable instruments on detection throughout the preferred programming languages.”

Each of the quoted numbers are mirrored on the SWE-Bench web site, however there are two points. Neither quantity has as but been verified by SWE-Bench, and the Amazon Q Developer rating on the Lite Leaderboard has dropped to #3. As well as, if there’s a supporting doc on the net for Amazon’s safety scanning declare, it has evaded my searches.

SWE-Bench, from Cornell, is “an analysis framework consisting of two,294 software program engineering issues drawn from actual GitHub points and corresponding pull requests throughout 12 standard Python repositories.” The scores replicate the answer charges. The Lite knowledge set is a subset of 300 GitHub points.

Let’s discover how Amazon Q Developer behaves on the assorted duties it helps in among the 15 programming languages it helps. This isn’t a proper benchmark, however slightly an try and get a really feel for a way properly it really works. Keep in mind that Amazon Q Developer is context delicate and tries to make use of the persona that it thinks most closely fits the setting the place you ask it for assist.

Predictive inline code technology with Amazon Q Developer

I attempted a softball query for predictive code technology and used one in all Amazon’s inline suggestion examples. The Python immediate equipped was # Perform to add a file to an S3 bucket. Urgent Possibility-C as instructed received me the code beneath the immediate within the screenshot beneath, after an unlawful character that I needed to delete. I needed to sort import on the high to immediate Amazon Q to generate the imports for logging, boto3, and ClientError.

I additionally used Q Chat to inform me how you can resolve the imports; it recommended a pip command, however on my system that fastened the flawed Python setting (v 3.11). I needed to perform a little sleuthing within the Frameworks listing tree to remind myself to make use of pip3 to focus on my present Python v 3.12 setting. I felt like singing “Daisy, Daisy” to Dave and complaining that my thoughts was going.

I additionally tried Amazon’s two different built-in inline suggestion examples. The instance to finish an array of faux customers in Python principally labored; I had so as to add the closing ] myself. The instance to generate unit checks failed after I pressed Possibility-C: It generated unlawful characters as a substitute of operate calls. (I’m beginning to suspect a problem with Possibility-C in VS Code on macOS. It might or might not have something to do with Amazon Q Developer.)

After I restarted VS Code, tried once more, and this time pressed Return on the road beneath the remark, it labored effective, producing the test_sum operate beneath.

# Write a check case for the above operate.
def test_sum():
    """
    Unit check for the sum operate.
    """
    assert sum(1, 2) == 3
    assert sum(-1, 2) == 1
    assert sum(0, 0) == 0

AWS exhibits examples of completion with Amazon Q Developer in as much as half a dozen programming languages in its documentation. The examples, just like the Python ones we’ve mentioned, are both quite simple, e.g. add two numbers, or relate to widespread AWS operations supported by APIs, comparable to importing information to an S3 bucket.

Pure language to code technology with Amazon Q Developer

Since I now believed that Amazon Q Developer can generate Python, particularly for its personal check examples, I attempted one thing a bit of completely different. As proven within the screenshot beneath, I created a file known as quicksort.cpp, then typed an preliminary remark:

//operate to type a vector of generics in reminiscence utilizing the quicksort algorithm

Amazon Q Developer saved making an attempt to autocomplete this remark, and in some circumstances the implementation as properly, for various issues. However it was simple to maintain typing my specification whereas Amazon Q Developer erased what it had generated, and Amazon Q Developer finally generated a virtually right implementation.

Quicksort is a well known algorithm. Each the C and C++ libraries have implementations of it, however they don’t use generics. As an alternative, it’s essential to write type-specific comparability features to move to qsort. That’s historic, because the libraries have been applied earlier than generics have been added to the languages.