Trapped Ion

IonQ

All-to-all connectivity. Commercial trapped-ion quantum computing with industry-leading algorithmic qubit performance.

  • #AQ 36 (Forte)
  • 99.5%+ gate fidelity
  • All-to-all connectivity
  • 4 cloud providers

About IonQ

Trapped ions, trapped in nothing else

IonQ was founded in 2015 by Christopher Monroe and Jungsang Kim, both leading researchers in trapped-ion quantum information at the University of Maryland and Duke University. The company went public in 2021, becoming the first pure-play quantum computing company listed on the NYSE.

IonQ uses Ytterbium-171 ions suspended in a Paul trap using oscillating electric fields. Each ion is an identical copy of every other: unlike superconducting qubits, which are fabricated and vary in frequency, every IonQ qubit is a single atom with precisely known quantum states. Laser pulses drive transitions between the ground state |0> and the excited state |1>. Two-qubit gates exploit the shared motional modes of the ion chain, which means any two qubits can interact directly without routing through intermediate qubits.

IonQ introduced the algorithmic qubit (#AQ) metric to quantify the useful computational capacity of a quantum processor, accounting for both qubit count and gate fidelity. A system with #AQ N can reliably run a random circuit of depth N using N qubits. This is more useful than raw qubit count alone.

Hardware Specifications

System specs at a glance

Specification Value
Current systems IonQ Aria, Forte, Forte Enterprise; Tempo rolling out
Algorithmic qubits (#AQ) #AQ 25 (Aria), #AQ 36 (Forte)
Two-qubit gate fidelity 99.5%+
Qubit technology Ytterbium trapped ions
Connectivity All-to-all (any qubit pair)
Coherence time Seconds to minutes
Gate time (2Q) ~100–600 microseconds
Cloud access AWS Braket, Azure Quantum, Google Cloud, IonQ Cloud
Supported SDKs Qiskit, Cirq, Amazon Braket SDK, Q#, PennyLane

  • IonQ Aria

    #AQ 25

    25 algorithmic qubits with 99.5%+ two-qubit gate fidelity. Available via Azure Quantum and IonQ Cloud (no longer listed on AWS Braket). Best for mid-depth circuits and learning.

  • IonQ Forte

    #AQ 36

    IonQ's flagship system with 36 algorithmic qubits. Forte introduced acousto-optic deflector (AOD) beam steering for software-configurable individual qubit addressing. Higher #AQ enables deeper, more complex algorithms. Available via AWS Braket, Azure Quantum, and IonQ Cloud.

Use Cases

Where IonQ hardware excels

  • Deep variational circuits

    All-to-all connectivity means QAOA and VQE circuits compile with zero SWAP overhead, preserving circuit depth and fidelity.

  • Quantum error correction research

    Long coherence times (seconds) combined with high fidelity gates enable multi-round error correction experiments.

  • Quantum chemistry simulations

    High gate fidelity reduces accumulated error in deep fermionic simulations, making molecular energy calculations more accurate.

  • Optimization problems

    Native connectivity maps naturally to fully-connected Ising models used in combinatorial optimization without routing penalties.

  • Quantum machine learning

    Low noise and high fidelity improve gradient estimates in parameterized quantum circuits used for QML experiments.

  • Benchmarking and research

    The #AQ metric provides a hardware-agnostic way to assess true algorithmic throughput, useful for comparing platforms.

Getting Started

Run your first circuit on IonQ

  1. Create an AWS account (easiest path)

    The fastest way to access IonQ hardware is via Amazon Braket. Create a free AWS account at aws.amazon.com. New accounts receive AWS Free Tier credits. Braket simulators are free within the first hour per month.

  2. Install the Amazon Braket SDK

    pip install amazon-braket-sdk

    The SDK supports Python 3.9+. Set your AWS credentials via aws configure or environment variables. See the Braket reference for full setup details.

  3. Write and submit your first circuit

    from braket.aws import AwsDevice
from braket.circuits import Circuit

device = AwsDevice("arn:aws:braket:us-east-1::device/qpu/ionq/Forte-1")

bell = Circuit().h(0).cnot(0, 1)
task = device.run(bell, shots=100)
print(task.result().measurement_counts)

    This creates a Bell state and runs it on IonQ Forte. With all-to-all connectivity, the CNOT compiles directly with no SWAP insertions needed.

  4. Use Cirq for Google Cloud access

    If you prefer Cirq, install cirq-ionq and authenticate with an IonQ API key. IonQ hardware can also be reached through the Google Cloud Marketplace listing. See the Cirq reference and the Cirq hello world tutorial. 

    pip install cirq-ionq

  5. Apply for research credits

    Academic researchers and students can apply for free QPU time via the IonQ Research Program. Credits are awarded based on research merit.

Pricing

Access costs across providers

IonQ hardware is available through multiple cloud providers, each with different pricing models. Costs are incurred per shot (measurement) and sometimes per task (job submission).

  • AWS Braket

    Per shot + per task

    $0.30 per task plus $0.08 per shot on IonQ Forte (per the AWS Braket pricing page). Free Tier: 1 hour of simulator time per month during your first 12 months (simulator only).

    View pricing details →

  • Azure Quantum

    Credits + pay-per-use

    First-time users receive $500 USD in Azure Quantum Credits per participating provider. IonQ hardware is billed in Azure Quantum Tokens, calculated from gate-shots, after credits are exhausted.

    View pricing details →

  • Google Cloud

    Pay per shot

    Available via Google Cloud Marketplace and billed through your Google Cloud account. See the marketplace listing for current rates. Requires Google Cloud billing account.

    View pricing details →

  • IonQ Cloud

    Free credits for researchers

    IonQ offers free QPU credits for academic researchers and students via the IonQ Research Program. Apply directly.

    View pricing details →

Learning Resources

Tutorials and reference docs