anyone trying to get into quantum computing or use it to build something there are plenty of SDKs available. Every large company and new startup work to create a Python package that can be used to learn the basics of quantum computing or develop an algorithm. But as someone who likes both software and quantum, I like to install a lot of packages just to try them out and advise people which one to use for what they want to do.
You search for a tutorial, enter something, then something else. Before you know it, you've got Qiskit, Cirq, PennyLane, and maybe something obscure sitting around your neighborhood… and no real idea of which one you should actually use to build whatever app you had in mind when you started. I am writing this article to save time.
The truth is (as you may already know), you don't need them all. You probably don't even need two.
The trick isn't choosing the “best” SDK (which I'd argue doesn't exist, yet), it's choosing the right one for what you're trying to do.
Quantum SDKs are immutable; they are built on very different priorities: some are education-first, some are hardware-first, some are machine-learning-first, and some are research playgrounds.
Once you realize that and have a clear goal in mind, the ecosystem becomes much less confusing. In this article, we'll discuss 4 of the most commonly used SDKs and generally talk about other tools that may not be commonly used.
So, let's get into it…
Qiskit: The “Default” Start Point
If you're not sure where to start, start here. If you Google the phrase “quantum computing,” you will find Qiskit on the first page of search results, if not the top 3 results!
Qiskit has become the de facto gateway for many people because it does several things well! It offers clean learning resources, access to real quantum hardware, and, most importantly, a large and active ecosystem.
Qiskit sounds like a neat place. He describes circuits, runs them, and gets results in a way that reflects how most people are taught quantum computing. And in many cases, computer science.
To create a simple circuit in Qiskit, you only need a few lines of code:
And just like that, you created a superposition and scaled it.
While Qiskit is great for teaching and learning, real-world testing, and general cycle-based workflows, it still struggles in many ways.
Because the purpose of building Qiskit is to provide a general framework for quantum computing, it is very general and vague, and sometimes poorly written, especially for new, application-specific tasks. Therefore, it can feel heavy, and is not ideal for gradient-based or ML-based workflows.
PennyLane: Built for Quantum Machine Learning
Now let's switch gears. If your interest in quantum computing includes enhancements, gradients, or machine learning, PennyLane is in a different category.
It is not designed as a general purpose SDK. It is specially designed for a combined quantum-classical workflow.
Most of the closest quantum algorithms look like this:
- Configure a parameterized quantum circuit.
- Run it.
- Measure something.
- Feed that to the classic optimizer.
- Repeat as needed to get the results you want.
This loop is where PennyLane shines.
For example, let's consider the main idea behind quantum machine learning: You have a quantum circuit with parameters, and you can optimize it using gradients. You can do that on PennyLane like this:
That last line is the key to why using PennyLane makes things so much easier here.
So, if you want to build an application that involves quantum machine learning, various algorithms, or gradient-based optimization, PennyLane is the way to go!
Now, you may have noticed that PennyLane's code is not as straightforward and precise as Qiskit; that is one of the problems of using it. It may have a slightly higher barrier to entry, and is less hardware-centric than Qiskit.
Cirq: For When You Want More Control
So far, we have Qiskit for general applications, PennyLane for QML, and now we have Cirq. Cirq sits in an interesting spot.
It's not as friendly as Qiskit, and it's not as ML-focused as PennyLane. Instead, it gives you low-level control over circuits and operations. This makes it popular in algorithm development, research, and hardware-aware circuit design.
Cirq feels closer to “metal.” He thinks more clearly about qubits, gates, and scheduling.
For example, I can replicate the same circuit built earlier with Qiskit, using Cirq like this:
For me, Cirq is often used when you want granular control over your circuit, need to maintain a research workflow, or want clean circuit descriptions. In turn, because it offers this hardware control, it has a steeper learning curve and a smoother ride than Qiskit and PennyLane.
It's worth noting that you can achieve the same level of hardware control with Qiskit as Cirq, with a shallower learning curve.
Amazon Braket: A Multi-Hardware Playground
Finally, let's talk about Braket! Brake is less about how you write circuits and more about what you use them for. It gives you access to quantum hardware providers through a single interface.
Different hardware platforms behave differently. Today, we have different forms of qubits: superconducting qubits, trapped ions, and photonic qubits, to name a few. Braket allows you to test these things without changing the ecosystem.
So, if you want cross-vendor hardware access, cloud-based workflows, or running tests, Braket is your tool of choice. That said, Braket is less intuitive than other SDKs and requires more context to be used effectively.
What many people forget
This is where things get fun and small! So, let's see examples of different approaches to quantum technology.
ID-Wave Ocean
So far, we have only discussed quantum computing at the gate! D-Wave's SDK is designed for quantum computing, not gateway computing. That means: you don't build circuits, but you define development problems.
Strawberry Fields: Photonic Quantum Computing
Most structures take qubits. Strawberry Fields doesn't want to. It works with continuous variable quantum systems, which are commonly used in photonic quantum computing.
qBraid: Closing the Ecosystem
qBraid tries to solve a real problem: What if you didn't have to choose just one SDK?
It allows you to convert circuits between frameworks such as Qiskit, Cirq, and others.
QuTiP, ProjectQ, and others
These tools tend to be advanced, highly specialized, and less user friendly.
They are useful, but not where you should start.
| If your goal | Start with… |
| Basics of learning | Qiskit |
| It works with real hardware | Qiskit |
| Quantum machine learning | Penny Lane |
| Research / algorithm control | Circ |
| Tries multiple hardware backends | Research/control algorithm |
So which one should you use?
To answer that question, you first need to know what you are trying to build! Let's say you want to build a variational circuit.
If you use Qiskit, you will manage the parameters manually, and optimization happens out of the box.
In PennyLane, parameters and gradients are built-in, and it's a natural ML workflow.
When you decide to use Cirq, you gain flexibility and take on more responsibility.
Same idea. A completely different experience. That's why the choice of SDK is important.
If you're still not sure, here's a simple tip:
- Start with Qiskit if you are new.
- Start with PennyLane if you care about ML.
- Move to Cirq if you want more control.
And don't touch anything else yet.
A final thought
I know all too well how confusing and frustrating it can be to decide which tool to use, let alone if you're new to the field! Things spiral out of control very quickly. The quantum ecosystem is not confusing because it is not well designed. It's confusing because it's still happening. Different tools exist because people are solving different problems using quantum computing.
So the question here is not “Which SDK is the best?”; technically, none of them are meant to be. Instead, the question you should ask yourself to help you decide is “Which SDK fits what I want to do?” Once you have the answer, everything becomes very clear. And more importantly, you can start building.
