Think

Ahura Quantum

The first fully integrated quantum environment for agent workflows.

Ahura is the first platform to put quantum computing inside the agent loop. Circuits, simulators and real quantum hardware sit behind the same MCP endpoint as memory, storage, inference and compute, so quantum becomes one more tool call.

An agent submits a circuit or problem, chooses the right machine for the job, and results land in its partition for the next step. No provider accounts, no separate toolchain.

What it does

Native to the agent loop
Quantum runs as a tool call alongside Memory, Store, AI, VM and Workflows. Nothing to wire up.
One interface, every machine
Simulators and quantum hardware from multiple providers are reached through the same tools.
Async with events
Long jobs finish in the background and announce completion.
Results stored for you
Output lands in your partition, ready for the next step.

An agent calls it like this

{ "tool": "qc.run", "circuit": { "openqasm": "…" }, "device": "local", "shots": 1000, "result": "counts" }

Interactive simulation

Try Ahura Quantum

Press Run to send the call. The dashboard shows what it would use in a partition. Simulated: nothing is actually executed.

ahura mcp · ptn_demo_7f3a

# Check a Bell-pair circuit.

→ POST /mcp

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "qc.validate",
    "arguments": {
      "circuit": {
        "openqasm": "OPENQASM 3; qubit[2] q; h q[0]; cx q[0], q[1];"
      },
      "device": "local"
    }
  }
}

# Run it on the local simulator.

→ POST /mcp

{
  "jsonrpc": "2.0",
  "id": 2,
  "method": "tools/call",
  "params": {
    "name": "qc.run",
    "arguments": {
      "circuit": {
        "openqasm": "…"
      },
      "device": "local",
      "shots": 1000,
      "result": "counts"
    }
  }
}

# Look up the task.

→ POST /mcp

{
  "jsonrpc": "2.0",
  "id": 3,
  "method": "tools/call",
  "params": {
    "name": "qc.task.get",
    "arguments": {
      "task": "qt_3c91"
    }
  }
}

Ready. Press Run.

ptn_demo_7f3a

Partition · N. Virginia · created just now

Simulated

Activeus-eastRetention hold

days left

Run time

0.0s

0 calls

Size

2.3MB

0.00 ACU used

Services in this partition

Quantum
0 calls · 0.00 ACU

Spend by service

Ledger

No usage yet.

Why it matters

Where quantum beats classical.

Quantum computers are not faster at everything. They are dramatically better at a narrow set of problems, and those problems matter.

Classical computing

Bits are 0 or 1. Each added bit adds one more bit of state.
Simulating a molecule exactly means tracking a state that grows exponentially with its electrons, so exact methods stall quickly.
Hard optimisation is tackled with heuristics that find good answers, not provably best ones.
Excellent, cheap and reliable for nearly everything else.

Quantum computing

Qubits hold superposition, so n qubits work across 2^n states at once.
Entanglement and interference let an algorithm amplify correct answers and cancel wrong ones.
Quantum systems are natural at simulating other quantum systems, which is what chemistry and materials are.
Known speedups are specific: some exponential, such as factoring and quantum simulation, some quadratic, such as unstructured search.

Chemistry and drug discovery
Model molecular behaviour directly instead of approximating it.
Materials science
Explore batteries, catalysts and superconductors at the quantum level.
Optimisation
Scheduling, routing and portfolio problems with enormous search spaces.
Cryptography and sampling
Tasks where quantum methods change what is computable or how fast.

Today’s quantum hardware is small and noisy, and for most practical problems it does not yet outperform the best classical methods. That is why Ahura Quantum pairs simulators with hybrid workflows: the quantum step is one tool call inside a larger classical process, so you gain the advantage where it exists and lose nothing where it does not.

In depth

What Ahura Quantum can do.

Two ways to describe a circuit
Send OpenQASM 3 (up to 64 KB) or a simple JSON gate list covering the standard gate set.
Several result types
Measurement counts, probabilities, full state vectors for small circuits, or expectation values for observables you define.
Validate first
Check a circuit, see its depth and gate counts, get a cost estimate for each machine and a suggested device before you spend anything.
Instant or background
Small circuits run immediately and return inline. Larger ones run in the background and announce completion.
Parameter sweeps
Submit up to 20 circuits at once, or use named parameters to vary one circuit.
Results kept for you
Outputs land in your partition and can be analysed later: marginal counts, fidelity against an ideal, expectations from counts.
Cost guards
A daily spending cap per account stops a loop from running up a large bill.
Hardware stays opt-in
Real quantum processors are switched off until you enable them.

Compute classes

Different machines for different jobs.

There is no single quantum computer. Hardware comes in several families with different strengths, so Ahura Quantum lists each one and what it is good at. The agent chooses the device.

Available runs todayOpt-in reachable through the same tools, switched on per accountComing soon planned, no date committed

Available quantum accelerators

Simulators and quantum hardware from leading providers, all behind one set of tools. The agent picks the machine that fits the job.

Local simulator
Available
Fast development simulator
Profile: Up to 20 qubits, 10,000 shots
Best for: Testing circuits instantly and cheaply before spending on bigger runs.
State-vector simulator
Available
Exact state-vector simulation
Profile: Up to 34 qubits
Best for: Validating algorithms exactly, including chemistry and optimisation circuits of moderate size.
Density-matrix simulator
Available
Noise-aware simulation
Profile: Up to 17 qubits
Best for: Seeing how a circuit behaves under realistic noise before it ever touches hardware.
Tensor-network simulator
Available
Structured-circuit simulation
Profile: Up to 50 qubits, limited circuit shapes
Best for: Large circuits with low entanglement structure that exact methods cannot reach.
IonQ Aria 1
Available
Trapped ion quantum computer
Profile: Up to 25 qubits, high-fidelity all-to-all connectivity.
Best for: Running complex algorithmic circuits with minimal gate routing overhead.
IonQ Forte 1
Available
Software-configurable trapped ion QPU
Profile: Up to 36 qubits, dynamic optical addressing.
Best for: Executing dense quantum algorithms requiring higher qubit counts and flexible gate configurations.
Rigetti Ankaa-2
Available
Superconducting qubit processor
Profile: Up to 84 qubits, 2D square lattice topology.
Best for: High-throughput gate-based execution and hybrid quantum-classical algorithm testing.
QuEra Aquila
Available
Neutral-atom analog quantum processor
Profile: Up to 256 qubits (rubidium atoms), field-programmable spatial layout.
Best for: Analog Hamiltonian Simulation (AHS), quantum optimization, and many-body physics problems.
OQC Lucy
Available
Superconducting qubit processor
Profile: Up to 8 qubits, ring architecture.
Best for: Low-latency execution on European data sovereignty hardware endpoints.
Superconducting gate-model
Opt-in
Circuits of superconducting loops cooled to near absolute zero
Profile: The fastest gates, measured in nanoseconds. Limited connectivity between qubits and short coherence, so circuits stay shallow.
Best for: Variational and hybrid algorithms, optimisation heuristics and near-term chemistry experiments where speed matters.
Trapped-ion gate-model
Opt-in
Individual charged atoms held in electromagnetic traps
Profile: The highest gate fidelities and long coherence, with every qubit able to talk to every other. Gates are slower and qubit counts are smaller.
Best for: Deeper, precision-sensitive circuits where accuracy matters more than speed, such as chemistry and error-correction research.
Neutral-atom analog
Opt-in
Atoms arranged and moved by laser tweezers, evolved as a whole
Profile: Large qubit counts and reconfigurable geometry. Runs analog Hamiltonian simulation rather than gate circuits.
Best for: Simulating quantum materials and magnetism, and graph-structured optimisation problems.
Neutral-atom gate-model
Coming soon
Laser-tweezer atom arrays running gate circuits
Profile: Hundreds to thousands of qubits with flexible connectivity. Gate fidelity and speed trail the best trapped-ion and superconducting machines.
Best for: Large problems that benefit from many qubits and a connectivity that can be reshaped mid-computation.
Quantum annealing
Coming soon
Purpose-built analog hardware that settles into low-energy states
Profile: Not general-purpose. Targets optimisation problems written as QUBO or Ising models.
Best for: Large combinatorial optimisation such as scheduling, routing and portfolio selection.
Photonic
Coming soon
Particles of light steered through optical circuits
Profile: Qubits that resist decoherence and can travel over fibre. Making photons interact is the hard part, and the field is earlier in its development.
Best for: Sampling problems, quantum networking and, in time, modular, room-temperature machines.
Silicon spin
Coming soon
Electron spins in silicon, built with chip-making processes
Profile: The promise of scale through existing semiconductor manufacturing. Still early, with small processors today.
Best for: Long-term scaling of general-purpose machines. Mostly of research interest for now.

Which machine for which job

Workload	Best-suited class	With Ahura
Optimisation (scheduling, routing, portfolios)	Annealing or gate-model QAOA	Simulators now. Superconducting or annealing hardware later.
Molecular energy and drug discovery	High-fidelity gate-model (VQE-style)	Simulators now. Trapped-ion and superconducting hardware opt-in.
Materials and magnetism simulation	Neutral-atom analog	Tensor-network simulation now. Neutral-atom hardware opt-in.
Algorithm development and teaching	Simulators	Available now. Start on the local simulator, scale up to the state-vector simulator.
Noise and error studies	Noise-aware simulation, then hardware	Density-matrix simulation now.
Large structured circuits	Tensor-network simulation	Available now.

Machine catalogues, qubit counts and availability change often and are set by the providers. This page describes the classes, not a guaranteed list of machines. Quantum hardware does not yet beat classical computing on most practical problems, which is why simulators and hybrid workflows come first.

Every call, listed

9 tools your agent can call. Each one is a standard MCP tool with a typed schema and a price shown in tools/list.

Run

qc.devices: Available machines with status, queue depth, limits and prices.
qc.validate: Check a circuit and estimate cost per device.
qc.run: Run a circuit and choose the result type.
qc.batch: Run up to 20 circuits.

Tasks

qc.task.get: Status, result and actual cost.
qc.task.list: Tasks by status or tag.
qc.task.wait: Wait up to 20 seconds for a task to finish.
qc.task.cancel: Cancel a queued or running task.

Analysis

qc.analyze: Marginals, fidelity and expectations from a finished task.

Use cases with Quantum

All use cases

Industries using Ahura Quantum

Works with

Next: Ahura GPU

Ahura AI

Cheap inference for everyday turns.

Ahura GPU

Scale-to-zero acceleration for inference and training.

Ahura Memory

Fast working memory for agents.

Ahura Store

Persistent files and objects.

Partitions

Where this service's data lives and how it freezes.

Ahura Quantum

What it does

An agent calls it like this

Try Ahura Quantum

Where quantum beats classical.

Classical computing

Quantum computing

What Ahura Quantum can do.

Two ways to describe a circuit

Several result types

Validate first

Instant or background

Parameter sweeps

Results kept for you

Cost guards

Hardware stays opt-in

Different machines for different jobs.

Available quantum accelerators

Local simulator

State-vector simulator

Density-matrix simulator

Tensor-network simulator

IonQ Aria 1

IonQ Forte 1

Rigetti Ankaa-2

QuEra Aquila

OQC Lucy

Superconducting gate-model

Trapped-ion gate-model

Neutral-atom analog

Neutral-atom gate-model

Quantum annealing

Photonic

Silicon spin

Which machine for which job

Every call, listed

Run

Tasks

Analysis

Use cases with Quantum

Industries using Ahura Quantum

Works with