Skip to Content
SynapseX ChatQuantum backends

See available quantum backends from Chat

SynapseX Chat is connected to a live QCOS instance, so you can ask which quantum suppliers and backends exist right now and get a real answer instead of a guess. This page shows you how to ask, how to verify the same list yourself with one command, and exactly where the boundary is: chat can design a quantum run, but it cannot submit one.

What you get

A current list of quantum suppliers reachable through the connected QCOS instance, which the assistant can read and explain. The live list includes a statevector simulator, LUMI HPC at CSC Finland, Azure Quantum, IBM Quantum, AWS Braket and several QCOS engines. Each entry carries whether it is configured and whether it is enabled, so you can tell an available backend from a listed one.

Ask in chat

  1. Keep the Quantum QCOS capability on in Settings > Capabilities. It is on by default.

  2. Make sure the lightning chip is Off for this turn — a Turbo turn runs without tools, and this question needs one.

  3. Ask:

    Which quantum backends are currently available through QCOS?
For each one, tell me whether it is a simulator or real hardware,
and whether it is configured.

The assistant calls the provider listing tool and answers from the response.

Verify it yourself

The provider listing is public, so you can check it directly:

curl -s https://chat.synapsex.ai/api/quantum/providers

The response looks like this (two entries out of a longer list, and each entry carries a few more fields than are shown here):

{
  "source": "qcos",
  "providers": [
    {
      "id": "statevector_sim",
      "name": "SoftQuantus Statevector Simulator",
      "configured": true,
      "enabled": true,
      "type": "simulator"
    },
    {
      "id": "azure_quantum",
      "name": "Azure Quantum",
      "configured": true,
      "enabled": true,
      "type": "cloud_qpu"
    }
  ]
}

Read it like this:

  • source — qcos means the list came from the connected QCOS instance, not a static file.
  • type — simulator or cloud_qpu tells you whether it is classical simulation or real quantum hardware behind a cloud provider.
  • configured — whether credentials for that supplier are in place.
  • enabled — whether the supplier is switched on.

The honest boundary

Chat can plan a quantum run. It cannot execute one. The assistant can help you design a submission — provider, shot count, hybrid mode, observables — but the planning tool explicitly returns a “not executed” result with no measurement data. Submitting a quantum job is not available on chat.synapsex.ai today.

So a prompt like “design a 4-qubit VQE run on the statevector simulator with 2000 shots” gives you a well-formed plan you can act on elsewhere. A prompt like “run it and show me the counts” will not produce counts.

Where you can actually run something today

The SynapseX CLI ships a free local quantum simulator that runs offline, on your own machine, with no account and no cost. A Bell state takes one command and returns real sampled measurement counts.

See The CLI quantum simulator for the Bell-state example and the exact gate support.

AI agentsLimits and errors