What Is Quantum Teleportation? How It Works and What It Transfers

Quantum teleportation transfers the state of a quantum system to another system using shared entanglement and ordinary communication. In the standard qubit protocol, it consumes one entangled pair and sends two classical bits. The original state is no longer available at the sender, and the receiver cannot recover it before the classical message arrives.

What gets teleported?

The object being transferred is quantum information. A photon can carry a state in its polarisation; another photon can receive that state. The original photon does not disappear from one laboratory and materialise in another. Bennett and colleagues’ 1993 proposal describes a way to reconstruct an unknown state using entanglement and a classical message.

For a qubit, the state can be written as α|0⟩ + β|1⟩. The complex amplitudes α and β determine measurement probabilities and interference; their squared magnitudes add to one. Teleportation transfers that state without first finding out the amplitudes and writing them in an ordinary message.

This is a more demanding task than sending the result of a single measurement. A measured 0 or 1 is one outcome obtained from the state. It is not a full description of the quantum information that was present before measurement. Recreating that outcome alone would not reproduce how the original system responds to other measurements.

Why not measure the state and send its description?

An unknown quantum state is not a hidden classical file that a sufficiently careful instrument can read from one specimen. Measurements in different bases generally cannot all be performed on that specimen while preserving its original state. A single outcome therefore cannot supply an exact description of an arbitrary input.

The related no-cloning theorem says that a universal physical operation cannot make perfect independent copies of an arbitrary unknown quantum state. This does not forbid copying a known classical bit or preparing many systems in a state whose preparation recipe is known.

Consider a useful check. A machine can copy the distinguishable states |0⟩ and |1⟩ by correlating them with a blank target. Applied to a superposition, the same linear operation instead produces correlations between the two systems. Two correlated systems are not two independent replicas of the original superposition. The copying rule that works for that classical basis does not become a universal quantum copier.

Teleportation respects this limit. It moves the quantum information while consuming the input’s original state. It does not supply a second freely usable copy. Keeping this constraint in view explains why the measurement at the sender is an essential part of the protocol.

What resources does quantum teleportation need?

Alice and Bob, the conventional names for sender and receiver, begin with an entangled pair. Alice holds one member and Bob holds the other. Alice also has a third qubit carrying the state to be transferred. This separation between the input qubit and the resource pair is central to the standard protocol.

They also need a channel for ordinary information and the ability to perform suitable operations and measurements. Entanglement is a resource that has to be established physically. Distributing it can involve photons, connections between processors or intermediate nodes. The teleportation step does not create a free channel across arbitrary distances.

The pair is consumed in one ideal transfer. Alice’s input is consumed as well, in the sense that it no longer carries the original state. The information ends up in Bob’s qubit once the relevant correction is known. Our entanglement explainer supplies the background for why the shared pair contains correlations that separate classical descriptions cannot reproduce.

How does the teleportation protocol work?

Alice performs a joint measurement on the input qubit and her member of the entangled pair. This is called a Bell measurement because it distinguishes four joint states. Its outcome determines which of four related versions of the input state Bob has, conditional on that outcome.

She then sends Bob the outcome as two classical bits. Four possibilities require two bits: 00, 01, 10 or 11. The message labels a correction; it does not report a measurement of α and β. Bob applies the corresponding operation, recovering the input state in the ideal protocol.

The corrections can be expressed as doing nothing, applying a bit flip, applying a phase flip or applying both. Which bit labels which correction depends on the convention for the measurement and bit order. IBM’s worked teleportation module gives an explicit circuit and calculation.

A bit flip exchanges the basis states |0⟩ and |1⟩. A phase flip changes their relative sign. Both can matter even when one measurement basis gives identical probabilities, because a relative phase affects interference and measurements in other bases.

The standard teleportation protocol in four steps

Why two bits do not contain the whole quantum state

An arbitrary qubit has continuously varying amplitudes, so it would be misleading to imagine those amplitudes compressed into a two-bit file. In an ideal teleportation protocol with a maximally entangled resource pair, Alice’s four outcomes are equally probable regardless of the input state. Her message alone cannot tell an observer which state she was given.

The two bits work because Bob already holds the other half of a suitable quantum resource. They tell him how to use that resource after Alice’s measurement. Removing the entangled pair does not leave an alternative two-bit method for sending arbitrary quantum information.

An analogy can help only up to a point: the message acts like an instruction selecting a correction, rather than the payload itself. Unlike an ordinary locked file, however, the resource and input obey quantum measurement rules. That is why duplicating the classical instructions does not let any number of recipients reconstruct independent perfect copies.

Does quantum teleportation work faster than light?

Bob cannot recover a chosen, usable input state before Alice’s message reaches him. If he has no access to her measurement outcome, averaging over the four possible outcomes leaves his qubit in the same maximally mixed state for every input. No local measurement on it reveals a message selected by Alice.

For example, Alice cannot encode a deliberate choice between two input states and have Bob read that choice immediately from his qubit. The conditional states differ, but Bob lacks the information specifying which condition occurred. Once the classical bits arrive, he can identify or account for the required correction.

A correction can sometimes be tracked in software rather than applied immediately as a physical gate. That bookkeeping does not change the communication limit: the necessary information still has to arrive. Entanglement supplies quantum correlations; it does not remove the ordinary signal delay.

This is also why Bell’s theorem and faster-than-light communication are different questions. A violation of a Bell inequality tests limits on a class of explanations for correlations. It does not provide a controllable signal that can be read before a classical message arrives.

Has quantum teleportation actually been demonstrated?

Yes. In a 1997 experiment, Bouwmeester and colleagues demonstrated teleportation of a photon’s polarisation state using an entangled-photon resource and an appropriate joint measurement. The paper’s later arXiv upload does not change the experiment’s original publication date.

In 2017, Ren and colleagues demonstrated ground-to-satellite teleportation over an uplink distance reaching 1,400 kilometres. For six tested input states they reported an average fidelity of 0.80 ± 0.01. Photons distributed the entangled resource across the physical link; the protocol did not move an object instantaneously.

Fidelity quantifies how closely the output matches the target state. A persuasive experiment must compare its performance with an appropriate classical strategy for the tested inputs and account for how successful events were selected. A small demonstration does not automatically establish an arbitrary-state, deterministic network service.

The long-distance result is a historical example, not a claim about the latest distance record. It shows that teleportation is experimental physics, while preserving the distinction between a successful transfer, a success probability and the engineering required for repeated practical operation.

Why is teleportation useful for quantum networks?

A quantum network aims to move quantum information or establish useful entanglement between nodes. Teleportation lets a node transfer a state once an appropriate entangled resource and a classical channel are available. The difficult work includes distributing that resource, storing it and matching the timing of operations.

Quantum repeaters are intended to help establish long-distance entanglement through shorter segments. The foundational repeater proposal connects and improves imperfect entangled pairs rather than copying a passing unknown quantum state. Loss, noise, memory lifetimes and operation errors remain relevant to any implementation.

Entanglement swapping, used to connect segments, is related to teleportation: a joint measurement can establish entanglement between systems that were not directly prepared together. It still requires communication to know the outcome and use the resulting resource. “Quantum internet” therefore names a network capability, not an instant replacement for all ordinary internet traffic.

Why teleportation also appears inside quantum computers

Teleportation can do more than transfer an unchanged state. Gottesman and Chuang’s 1999 work showed how modified resource states and teleportation can implement operations for quantum computation. This connects state transfer to the design of fault-tolerant operations.

The attraction is that some of the work can be shifted into preparing a resource, followed by measurements and corrections. The costs then include how reliably that resource can be produced and how errors propagate. Our quantum-error-correction article explains why preserving information through those operations is a separate challenge.

The achievement is precise: quantum information can move between physical carriers without reading out a complete classical description of an unknown state. It requires entanglement, measurement and communication, obeys the no-cloning rule and respects the speed limit. Those conditions are what make teleportation a useful protocol and define what it can accomplish.