Four free, interactive tools that take QEC out of the papers and into your browser. No install. No login. Just the physics, made tangible.
A five-minute run-up before the tools — what a qubit is, why it breaks, and how error correction fights back. Skip it if you already live here.
A classical bit is 0 or 1. A qubit is a point on the Bloch sphere — any state |ψ⟩ = α|0⟩ + β|1⟩ with |α|² + |β|² = 1. The poles are the classical states; everywhere else is genuinely quantum.
That continuous freedom is the whole source of quantum advantage — and the whole reason error correction is hard. There's no discrete "flip" to simply re-measure and fix.
Superposition lets one qubit carry a weighted blend of |0⟩ and |1⟩ simultaneously. Entangle several qubits and their joint state can't be written as separate individual states at all — measuring one instantly constrains the others.
QEC codes are built entirely out of entanglement: a logical qubit is smeared redundantly across many physical qubits, so no single physical error can read out — or destroy — the encoded information.
Qubits leak into their environment constantly. Bit flips (X), phase flips (Z), and their combination (Y) accumulate at rates measured in microseconds — long before most useful algorithms finish running.
Worse: you can't just measure a qubit to check if it flipped. Measurement collapses the superposition itself. QEC has to detect errors without ever learning the state it's protecting.
Stabilizer codes measure joint parity operators (stabilizers) across groups of data qubits using ancilla qubits — without ever measuring the data qubits themselves. A flipped ancilla outcome (a "syndrome") flags that an error occurred nearby, without revealing what the logical state actually is.
Read syndromes round after round and you get a defect pattern in space and time — exactly what a decoder like MWPM is built to interpret.
Data and ancilla qubits alternate on a 2D lattice. Each ancilla stabilizes its neighboring data qubits; syndromes light up on a checkerboard pattern whenever an error occurs. It's local, planar, and matches real superconducting and neutral-atom hardware layouts — which is why it's the leading candidate for fault tolerance today.
Every tool in the suite below is a different lens on this same lattice: how errors move through it, how decoders read it, and how logical gates are performed on it.
Each tool isolates a core concept in quantum error correction — from noise thresholds to spacetime decoding to the MWPM algorithm to lattice surgery to building stabilizer codes and applying logical gates by hand.
Errors in a quantum computer don't just happen in space — they persist through time. This tool renders the full 3D spacetime volume that modern decoders like Harmony and Sparse Blossom actually operate on. Inject errors, watch worldlines form, toggle decoder matching overlays.
Minimum Weight Perfect Matching is the algorithm running inside almost every serious surface code experiment today. This tool slows it down to human speed. Inject X errors, watch the defect graph form, see every matching evaluated, and get a verdict: corrected or logical error.
The fault-tolerance threshold is the single most important number in QEC. Drag the noise slider below it and watch logical error rates collapse exponentially with each concatenation level. Push above it and watch them explode. The threshold theorem, made visceral.
Transversal gates are a luxury real hardware doesn't have. Google and IBM run their two-qubit logical gates by merging and splitting surface-code patches along their boundaries. Pick initial states for two patches, choose a merge type, and step through the full protocol.
Pick a stabilizer code — 3-qubit repetition codes or the [[7,1,3]] Steane code — and build it from scratch. Inject Pauli errors on any physical qubit, watch exactly which generators fire, and let the tool propose the minimum-weight correction your syndrome implies.
Watch H, S, and CNOT act transversally on the Steane code — one physical gate per physical qubit, no error spreading. Then see why T can't: step through gate teleportation via a magic state, the standard way stabilizer codes get a universal gate set past the Eastin–Knill theorem.
Focused on surface codes, decoding architectures, and fault-tolerant systems.
QEC is one of the most consequential problems in science right now — and almost none of it was visible outside a handful of research groups. That felt wrong.
These four tools exist to close that gap: free, interactive, and built directly from the algorithms running in real labs today.