Run a quantum program.
Get an answer you can check.
Write OpenQASM, or start from an example. The service picks the strongest way to simulate your circuit, tells you which and why, says how exact the answer is, and signs it.
Result
readyRun a program to see its result here.
Counts are keyed by the program's classical bits, highest bit first. With no measurement, every qubit is read into the bit of its own index.
A receipt shows that one key signed this request and this result; anyone can run the request again to compare. It does not say who holds the key. wai-quantum verify checks a downloaded job.
From an agent
Any client that speaks the Model Context Protocol can use the service as tools: add one URL, with no key. The tools run a circuit, read a job, verify a receipt, describe the service, and list worked examples. Each answer comes as a short summary written for a model to read, with the whole result alongside.
{
"mcpServers": {
"wai-quantum": {
"type": "http",
"url": "https://wai.transaction.science/quantum-jobs/mcp"
}
}
} For other programs: the OpenAPI description, the examples as JSON, and a summary for language models at /llms.txt.
What it can run
auto tries these in turn and takes the first that fits the circuit and what you asked for. The limits are
the service's own, read live.
| Method | Takes | Up to | Accuracy |
|---|
To build circuits by hand and watch them evolve, use the quantum IDE, which runs in the tab. To learn the ideas, take the course.