Compare Quantum SDKs
A vendor-neutral side-by-side of the major software development kits — because the right tool depends on your goal, not on whose brand you saw first.
| SDK | Vendor | Language | Hardware access | Open source | Best for |
|---|---|---|---|---|---|
| Qiskit | IBM | Python | IBM Quantum (superconducting) | General-purpose circuits, the largest ecosystem and learning material, real IBM hardware access. | |
| Cirq | Python | Google (research) + simulators | Fine-grained control over gates, timing, and noise; NISQ experiments. | ||
| PennyLane | Xanadu | Python | Hardware-agnostic (plugins for many backends) | Quantum machine learning and differentiable/variational circuits; autodiff across backends. | |
| Amazon Braket SDK | AWS | Python | IonQ, Rigetti, IQM, QuEra via AWS | Running the same code across multiple vendors' hardware from one cloud account. | |
| Q# / QDK | Microsoft | Q# (+ Python interop) | Azure Quantum providers | A dedicated quantum language with strong typing; algorithm-focused, resource estimation. | |
| CUDA-Q | NVIDIA | C++ / Python | GPU simulation + QPU backends | GPU-accelerated simulation and hybrid quantum-classical / HPC workloads at scale. | |
| TKET (pytket) | Quantinuum | Python (C++ core) | Many backends (compiler/router) | Best-in-class circuit optimization and qubit routing; retargeting circuits to real device topologies. | |
| Stim | Craig Gidney / Google | Python / C++ | Stabilizer simulation only | Blazing-fast stabilizer + error-correction simulation; the standard tool for QEC research. |
IBM · Python · open source
Hardware: IBM Quantum (superconducting)
Best for: General-purpose circuits, the largest ecosystem and learning material, real IBM hardware access.
Google · Python · open source
Hardware: Google (research) + simulators
Best for: Fine-grained control over gates, timing, and noise; NISQ experiments.
Xanadu · Python · open source
Hardware: Hardware-agnostic (plugins for many backends)
Best for: Quantum machine learning and differentiable/variational circuits; autodiff across backends.
AWS · Python · open source
Hardware: IonQ, Rigetti, IQM, QuEra via AWS
Best for: Running the same code across multiple vendors' hardware from one cloud account.
Microsoft · Q# (+ Python interop) · open source
Hardware: Azure Quantum providers
Best for: A dedicated quantum language with strong typing; algorithm-focused, resource estimation.
NVIDIA · C++ / Python · open source
Hardware: GPU simulation + QPU backends
Best for: GPU-accelerated simulation and hybrid quantum-classical / HPC workloads at scale.
Quantinuum · Python (C++ core) · open source
Hardware: Many backends (compiler/router)
Best for: Best-in-class circuit optimization and qubit routing; retargeting circuits to real device topologies.
Craig Gidney / Google · Python / C++ · open source
Hardware: Stabilizer simulation only
Best for: Blazing-fast stabilizer + error-correction simulation; the standard tool for QEC research.
Which should a beginner pick?
Start with Qiskit (largest community and learning material) or PennyLane (if you're drawn to machine learning). Both are Python, free, and run on simulators without any hardware account. You can try circuits right now in the Quantum Sandbox — it even exports Qiskit code.