Quantum Circuit Simulator Online: What It Computes and How to Use One

What you are actually asking a simulator to do

A quantum circuit is a short program. It has a set of qubits, drawn as horizontal wires, and a sequence of gates placed on those wires from left to right. At the end, the qubits are measured and you get a string of zeros and ones.

A quantum circuit simulator is a classical program that works out what an ideal quantum device would do with that circuit. Nothing quantum happens inside it. It is linear algebra on an ordinary processor, which is exactly why it is a good place to study: there is no noise, no queue, and you can look at things a real device never shows you.

If you are searching for a quantum circuit simulator online, you most likely want one of two things: a place to try gates and see what they do, or a way to check your own reasoning about a circuit you met in a course or a textbook. Both come down to the same loop: predict, run, compare.

The state vector, in plain terms

The most common kind of simulator is a state-vector simulator. It keeps a list of complex numbers called amplitudes, one for every bitstring the qubits could show when measured. Two qubits have four possible bitstrings, so the list has four entries. Add a qubit and the list doubles.

Each gate is a rule for mixing those amplitudes:

• X swaps the amplitudes of zero and one on its wire. It is the quantum version of a NOT.
• H, the Hadamard gate, turns a definite zero into an equal-weight combination of zero and one. Applied twice, it undoes itself.
• Z, S and T leave the measurement probabilities alone and change only the phase of the one-amplitude, by different angles.
• Y combines a bit flip with a phase change.
• CNOT acts on a pair of wires: it flips the target only in the part of the state where the control is one. This is the gate that creates entanglement.

When the circuit ends, the probability of each bitstring is the squared magnitude of its amplitude. That rule, the Born rule, is the bridge from the hidden list of amplitudes to what you would see in a lab.

Probabilities versus counts

A simulator can show you two different things, and it helps to keep them apart.

Probabilities are exact. They come straight from the amplitudes and are the same every time you run the circuit.

Measurement counts imitate an experiment. On real hardware each run of the circuit, called a shot, gives one random bitstring, and you tally many shots. Some simulators reproduce this by drawing random samples, so the tallies shift a little from run to run. Others simply report the expected tally, the probability multiplied by the number of shots, which is the same every time. It is worth knowing which kind you are using.

The classic first example is the Bell pair: a Hadamard on the first qubit, then a CNOT from the first to the second. The exact probabilities are one half for both-zero and one half for both-one, and nothing for the mixed outcomes. On a device, or in a sampling simulator, the counts will be close to an even split but rarely exactly even. Either way the mixed outcomes never appear, and that missing half of the table is the quickest way to see what "the qubits are correlated" means, and what it does not mean.

What a histogram cannot tell you

Phase gates are where beginners most often get confused, because they appear to do nothing. Put a Hadamard on a qubit and you get an even split. Put a Hadamard and then a Z, and you get the same even split. The two states are different, but a plain measurement cannot tell them apart.

The difference shows up only when you let the amplitudes interfere, for example by adding a second Hadamard. One circuit then returns zero every time, the other returns one every time. A simulator is useful here precisely because it holds the full state: a tool that compares states, and not only histograms, can tell you that you built the wrong one even when the bars look right.

Where simulation stops

Because the amplitude list doubles with every added qubit, a state-vector simulator runs out of memory and time at a modest number of qubits. Every tool of this kind has a cap, whether it runs in your browser or on a server. That is not a defect of a particular product; it is the reason people build quantum hardware.

Two consequences for a learner:

• A simulator is for understanding circuits, not for getting answers faster than a classical computer. It is a classical computer.
• Ideal simulation leaves out noise. Real devices lose coherence and make gate errors, so results on hardware are messier than what you will see here.

What the QuantumNative Circuit Lab does

QuantumNative is a quantum computing education product, and its web app includes a Circuit Lab built for the predict-run-compare loop described above. Concretely:

• A click-to-place circuit grid. You pick a gate from a palette of H, X, Y, Z, CNOT, T and S, click a slot on a qubit wire to place it, and click a placed gate to remove it. You can add or remove wires up to your plan's limit.
• Runs on a state-vector simulator. Pressing run sends the circuit to QuantumNative's server, where a classical state-vector simulator computes the result. The lab does not run on quantum hardware and does not say it does.
• Both views of the result. The result panel shows the probability of each basis state as bars, and a count for each outcome. The counts are expected values calculated from the probabilities, not random samples, so repeating a run gives identical numbers.
• Missions that check the state, not the picture. A set of guided missions asks you to build a specific target: an equal superposition, a phase-flipped state, a Bell pair, a three-qubit GHZ state, and Deutsch's algorithm. After each run, the lab compares the simulated state with the target by computing their fidelity and tells you whether you reached it. The check ignores global phase, and for Deutsch's algorithm it also checks that the oracle and the surrounding Hadamards are present, so a shortcut to the same final state does not pass. Each verdict is labelled as checked on a classical simulator.
• An interference experiment. A separate lab page models a Mach-Zehnder interferometer: you open or block a path, predict where the photon is detected, then run it. The outcome is computed from an exact model and cross-checked against the same server simulator.

The lab also has a Bloch sphere view for single-qubit rotations; the Free plan limits how many gates you can apply there.

What it is not

To set expectations honestly:

• It is a teaching tool with a small, fixed gate palette, not a general-purpose research simulator. There are no rotation-angle gates or custom gates in the circuit grid.
• It requires a signed-in account. The Circuit Lab is part of the Free plan with a lower cap on qubit wires, and the Pro plan raises the cap; see pricing for the current plans.
• It shows ideal results only: no noise model and no shot-to-shot randomness in the circuit grid.
• Mission completion is recorded in your browser, not as a credential.
• Using it does not by itself make anyone proficient. It gives you a place to test predictions; the understanding comes from making them.

A good first session

If you have never used a circuit simulator, this sequence covers the ideas on this page:

• Place a single H and run it. Read the probabilities, then the counts, and work out how one follows from the other.
• Add a Z after the H. Notice that the result panel looks the same, then add a second H and see that it no longer does.
• Build the Bell pair with H and CNOT, and look at which outcomes are missing.

You can create an account and open the Circuit Lab from the dashboard to try it.