QFlow StudioLive Build workspace

The circuit and the code, in one working view.

Work on the Bell State circuit beside its generated Qiskit. Select an operation to trace it across both views, then drag the divider to give Design or Code more room.

Open Bell State Browse samples

Sample: Bell State
Circuit: 2 qubits
Views: Design + Code
Sync: Operation-level

Workflow / active draft

Ready

Bell State Demo

4,096 shots

Bell State workflow

Qiskit Aer / 4,096 shots

zoom 100%canvas

01from qiskit import QuantumCircuit02qc = QuantumCircuit(2, 2)03qc.h(0)04qc.cx(0, 1)05qc.measure([0, 1], [0, 1])06result = qflow.run(qc, runner='qiskit-aer')

PythonQiskitsource synced

Compatibility layer

Your stack stays yours.

QFlow connects the circuit record to the SDKs, formats, and provider access your team already uses.

default limit: 31q
simulators: 04
source formats: 02
review URL: 01

Review integration status

runtimes & access

IBM Quantumruntime
AWS Braketaccess
Azure Quantumcloud
IonQtrapped ion
Rigettisuperconducting
Quantinuumpytket

SDKs & formats

QiskitSDK
CUDA-QNVIDIA hybrid
CirqGoogle circuits
OpenQASMexport

Names and marks belong to their owners. Supported integration context, not endorsement.

01 / Run record

Keep each circuit run in one place.

Store the brief, canvas, source, route, run status, trace, result, and share link together.

System of record

One circuit. One owner. Every decision attached.

linked stages

source formats

review URL

run record

QF-2048 supply-chain optimization

single source

Brief

objective + owner

Circuit

8 qubits / 24 ops

Source

Qiskit + QASM3

Runner

Aer / Cirq / qsim

Run

4,096 shots

Result

counts + trace

One live run record.

Version, owner, source, run state, and result move as one record.

versionv14

ownerQuantum Ops

runnerQiskit Aer

resultready

Controlled iteration

Each cycle starts with intent, stops for a human decision, and returns evidence to the same versioned record.

closed loop

01Intentbrief + owner
02Draftblocks + source
03Preflightroute + limits
04Human gateapprove submit
05Runcounts + trace
06Recordevidence + next

Recorded evidence becomes context for the next draft.

02 / Security boundary

Share results, not secrets.

Publish read-only result pages, certificate pages, and share-safe links while credentials and drafts remain private.

Review-ready

Counts, probabilities, source, and trace stay readable.

Share-safe

Public evidence omits provider credentials and private drafts.

Addressable

A stable result URL keeps the decision easy to revisit.

Inspect a complete evidence packet

Simulator result page

Bell-style counts

4,096 shots · probabilities · CSV/JSON export

complete

934

128

177

2857

Result page

counts, probabilities, JSON, CSV

Certificate page

issuer + completion status

Shared result

read-only public link

Private boundary

tokens stay sealed

03 / Technical depth

What actually runs today.

A circuit can move from canvas to source, runner, counts, exports, and share-safe evidence without leaving the workflow record.

31q

default cap

<=8q

statevector

off

raw Python

Can I run circuits here?

Yes

Visual blocks compile to CircuitOperation[] JSON, generated source, and runner requests.

Runner architecture

FastAPI + 4 simulators

The Docker runner executes Qiskit Aer, Cirq, qsim, bounded Stim, and OpenQASM 2/3 paths behind the Next.js app.

Qubit limit

31q default

Simulator runs accept up to 31 qubits by default; larger requests are rejected before execution.

State output

<=8q statevector

Small Qiskit runs return amplitudes; larger circuits keep counts and probabilities tractable.

Runtime boundaries

Power without an ambiguous execution surface.

current contract

6 noise modes

Noise models

Starters + workbench

VQE / QAOA

IonQ / IBM Eagle / Heron

Transpile targets

Off by default

Raw Python

01 Visual blocks

Circuit JSON

02 Source

Qiskit / OpenQASM

03 Runner

FastAPI runner

04 Evidence

Counts / LaTeX / CSV

04 / AI Builder

Review the exact circuit change where it will run.

QFlow marks the proposed gates on their qubit lanes and the matching generated source lines inside the same Build workspace. Apply and Run remain separate, explicit team decisions.

Bounded change

Two operations. Two source lines. One workflow record.

Route: unchanged
Parser: 0 warnings
Run: manual

BuildHub

QAOA supply-chain route

workflow / qaoa-route-v3

proposal pending

Assistant edit · proposal 014

review

Add RZ(q1, 0.42) and RX(q2, π/4)

2 circuit operations · 2 generated lines · measurement order and hardware route unchanged

Design/3q · compact circuit

diff

GatesFitAuto layout

Canvas 100%

savedproposed3q · 5 saved + 2 proposed

Code/Qiskit 2.x · generated

diff

workflow.qiskit.pyopenqasm.qasm

1+from math import pi2+from qiskit import QuantumCircuit3+ 4+qc = QuantumCircuit(3, 3)5+qc.h(0)6+qc.cx(0, 1)7+qc.rz(0.42, 1)8+qc.rx(pi / 4, 2)9+qc.measure([0, 1, 2], [0, 1, 2])

diff previewround-trip clean · 0 warnings

ResultsNo run · review and apply the proposed edit first

05 / Provider status

Check providers before you submit.

QFlow shows credentialed device context, route fit, SDK output, OpenQASM export, and result artifacts for the same circuit. Live provider status appears after credentialed sign-in.

Route before run

Provider choice becomes a reviewable decision, not a hidden script setting.

See live integration boundaries

provider fabric

credentialed access layer

QOS / PREFLIGHT

route decision

Fit, credentials, fallback, evidence.

provider adapters

simulators

workflow record

Access

keys + devices

Route

fit + fallback

Run

QASM + counts

IBM Quantum

Backend list, job submit, polling, and result retrieval use connected IBM access.

credentialed

Public stats

Account-scoped queue and backend state stay inside the signed-in provider hub.

signed-in status

Local runner

Qiskit Aer, Cirq, qsim, Stim, and OpenQASM paths produce counts and evidence without provider tokens.

live

provider access

hardware, cloud, runtime

6 connected

IBM Quantum

runtime

AWS Braket

access

Azure Quantum

cloud

IonQ

trapped ion

Rigetti

superconducting

Quantinuum

pytket

artifact layer

SDKs, code, results

4 connected

Qiskit

SDK

CUDA-Q

NVIDIA hybrid

Cirq

Google circuits

OpenQASM

export

Credentials stay on the private side.

Only review-safe route context and result artifacts move into shared evidence.

06 / Who it is for

Use QFlow for research, provider pilots, and classes.

Research operations, provider rollout, and academy cohorts start from circuits, source, and simulator output instead of a blank notebook.

Explore solutions

Abstract quantum research testbed used for research operations and pilot planning

Research testbed01

Research ops

Build Bell, GHZ, QAOA, VQE, and custom circuit drafts, then run simulator checks before hardware time.

Provider fabric02

Provider rollout

Compare credentials, device capabilities, queue context, and routing constraints before submitting quantum jobs.

Abstract quantum learning progression used to represent academy cohort workflows

Learning path03

Academy cohorts

Give students runnable labs with source, counts, progress, comments, and certificates instead of passive notes.

07 / Research briefings

Long reads tied to real quantum surfaces.

Source-backed briefings that connect current research to QFlow circuits, providers, algorithms, and learning flows.

View blog

Abstract quantum academy progression from lesson panels to a runnable circuit

5 min

QFlow Studio / AI-generated visual

education

5 min read

QFlow begins a community test pilot with Quantum Hacettepe

The scoped pilot will test how visual circuits, synchronized code, simulation evidence, and guided learning hold up in real student-team workflows.

sections

sources

minutes

Read long analysis

4 min

QFlow Studio / AI-generated visual

education

4 min read

Quantum learning platform workflows 2026: lessons to proof

A practical SEO guide to quantum learning platforms that turn lessons, circuit practice, provider context, certificates, and review evidence into one path.

Read long analysis