Complete specifications for every major quantum processor from IBM, Google, Quantinuum, IonQ, Rigetti, Atom Computing, Microsoft, and more. Sorted by qubit count and including fidelity metrics, coherence times, gate speeds, connectivity topology, operating temperature, cloud access, and current operational status. Data sourced from manufacturer publications, peer-reviewed papers, and verified company disclosures.
| Processor | Manufacturer | Approach | Qubits | 1Q Fidelity | 2Q Fidelity | Coherence | Gate Speed | Cloud | Year | Status |
|---|---|---|---|---|---|---|---|---|---|---|
| Advantage | D-Wave Systems | superconducting | 5,627 | - | - | ~20 µs (annealing time ~20 µs) | N/A (quantum annealer) | Yes | 2020 | commercial |
| Advantage2 | D-Wave Systems | superconducting | 4,400 | - | - | ~100 µs | N/A (quantum annealer) | Yes | 2024 | commercial |
| QuEra 3,000+ qubit system | QuEra Computing | neutral-atom | 3,000 | 99.9% | 99.9% | ~2 s | ~1 µs (Rydberg gate) | No | 2026 | research |
| Kookaburra | IBM | superconducting | 1,386 | - | - | - | - | No | 2026 | research |
| 1,180-atom Array | Atom Computing | neutral-atom | 1,180 | 99.5% | 99.0% | ~40 s | ~1 µs (Rydberg gate) | No | 2023 | prototype |
| Condor | IBM | superconducting | 1,121 | 99.5% | 98.5% | ~90 µs | ~340 ns (CNOT) | No | 2023 | research |
| Flamingo | IBM | superconducting | 462 | 99.9% | 99.5% | 200 µs | ~60 ns | No | 2025 | prototype |
| Osprey | IBM | superconducting | 433 | 99.5% | 99.0% | ~90 µs | ~340 ns (CNOT) | No | 2022 | research |
| Aquila | QuEra Computing | neutral-atom | 256 | 99.5% | 99.5% | ~2 s | ~1 µs (Rydberg gate) | Yes | 2023 | commercial |
| Borealis | Xanadu | photonic | 216 | - | - | N/A (photonic) | ~100 ps (optical gates) | Yes | 2022 | commercial |
| Fresnel | Pasqal | neutral-atom | 200 | 99.5% | 99.0% | ~1 s | ~1 µs (Rydberg gate) | Yes | 2024 | commercial |
| Heron r2 | IBM | superconducting | 156 | 99.9% | 99.7% | 200 µs | ~60 ns (ECR gate) | Yes | 2024 | commercial |
| Heron r1 | IBM | superconducting | 133 | 99.8% | 99.5% | 150 µs | ~70 ns (ECR gate) | Yes | 2023 | commercial |
| Eagle | IBM | superconducting | 127 | 99.5% | 99.0% | 100 µs | ~340 ns (CNOT) | Yes | 2021 | commercial |
| Willow | Google Quantum AI | superconducting | 105 | 99.93% | 99.7% | ~100 µs | ~25 ns (single), ~32 ns (two-qubit) | No | 2024 | research |
| SQale | Infleqtion | neutral-atom | 100 | 99.5% | 99.0% | ~1 s | ~1 µs | No | 2024 | prototype |
| planqc Atom Processor | planqc | neutral-atom | 100 | 99.5% | 99.0% | ~1 s | ~1 µs | No | 2024 | prototype |
| Helios | Quantinuum | trapped-ion | 96 | 99.99% | 99.9% | >10 s | ~150 µs (two-qubit) | Yes | 2025 | commercial |
| Ankaa-2 | Rigetti Computing | superconducting | 84 | 99.5% | 99.0% | ~25 µs | ~80 ns (two-qubit iSWAP) | Yes | 2024 | commercial |
| Wukong | Origin Quantum | superconducting | 72 | 99.5% | 97.0% | ~30 µs | ~200 ns (two-qubit) | Yes | 2024 | commercial |
| Tempo | IonQ | trapped-ion | 64 | 99.99% | 99.7% | >1 s | ~150 µs (two-qubit) | Yes | 2025 | commercial |
| H2-1 | Quantinuum | trapped-ion | 56 | 99.99% | 99.8% | >10 s | ~200 µs (two-qubit) | Yes | 2023 | commercial |
| IQM 54-qubit Processor | IQM Quantum Computers | superconducting | 54 | 99.9% | 99.5% | ~40 µs | ~60 ns (two-qubit) | No | 2025 | prototype |
| Sycamore | Google Quantum AI | superconducting | 53 | 99.85% | 99.4% | ~20 µs | ~25 ns (single), ~32 ns (two-qubit) | No | 2019 | research |
| Forte | IonQ | trapped-ion | 36 | 99.97% | 99.5% | >1 s | ~200 µs (two-qubit) | Yes | 2023 | commercial |
| Ankaa-3 | Rigetti Computing | superconducting | 36 | 99.7% | 99.5% | ~30 µs | ~60 ns (two-qubit) | Yes | 2025 | commercial |
| Toshiko | Oxford Quantum Circuits | superconducting | 32 | 99.7% | 98.5% | ~30 µs | ~80 ns | Yes | 2024 | commercial |
| Aria | IonQ | trapped-ion | 25 | 99.96% | 99.4% | >1 s | ~250 µs (two-qubit) | Yes | 2022 | commercial |
| PINE | Alpine Quantum Technologies | trapped-ion | 24 | 99.97% | 99.5% | >1 s | ~200 µs (two-qubit) | Yes | 2023 | commercial |
| H1-1 | Quantinuum | trapped-ion | 20 | 99.99% | 99.8% | >10 s | ~200 µs (two-qubit) | Yes | 2021 | commercial |
| Garnet | IQM Quantum Computers | superconducting | 20 | 99.8% | 99.0% | ~30 µs | ~80 ns (two-qubit) | Yes | 2023 | commercial |
| QuiX Quantum 20-mode Processor | QuiX Quantum | photonic | 20 | - | - | N/A (photonic) | ~ps (photonic) | No | 2023 | commercial |
| Harmony | IonQ | trapped-ion | 11 | 99.5% | 96.0% | >1 s | ~300 µs (two-qubit) | Yes | 2020 | commercial |
| Majorana 1 | Microsoft | topological | 8 | - | - | Theoretically very long (topological protection) | - | No | 2025 | research |
| Lucy | Oxford Quantum Circuits | superconducting | 8 | 99.5% | 97.0% | ~25 µs | ~100 ns | Yes | 2022 | commercial |
| Ocelot | Amazon Web Services | cat-qubit | 7 | - | - | - | - | No | 2025 | research |
| Diamond Quantum Accelerator | Quantum Brilliance | nv-center | 5 | 99.0% | 95.0% | ~1 ms (room temp) | ~10 ns | No | 2023 | prototype |
| Triangulum | SpinQ Technology | spin | 3 | 97.0% | 95.0% | ~1 s | ~ms (NMR pulses) | No | 2022 | commercial |
| Gemini Mini | SpinQ Technology | spin | 2 | 97.0% | 95.0% | ~1 s | ~ms (NMR pulses) | No | 2021 | commercial |
| Aurora | Xanadu | photonic | N/A | - | - | N/A (photonic) | ~100 ps (optical gates) | No | 2025 | research |
Qubit count alone is misleading. Fidelity and error correction matter more.
The processor with the most qubits is not necessarily the most capable quantum computer. Quantinuum's 56-qubit H2 system can run deeper, more complex circuits than many processors with 10 times the qubit count due to its 99.9975% two-qubit gate fidelity. Google's 105-qubit Willow demonstrated below-threshold error correction that larger, noisier processors cannot match. When comparing quantum computers, prioritize fidelity, error correction capability, and demonstrated algorithmic performance over raw qubit count.