> ## Documentation Index
> Fetch the complete documentation index at: https://docs.rhymeq.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Welcome to Rhyme

> A Python-based abstraction layer for quantum programming.

**Rhyme is a Python-based abstraction layer for quantum programming.** You write quantum programs with the values you already use in Python, such as numbers, strings, booleans and your own objects, instead of qubits, gates and circuits. Rhyme then compiles your program to CUDA-Q and runs it on real quantum hardware, not only on its built-in simulator.

```python theme={null}
from math import pi
from rhyme import qstr

# two possible moods at once
mood = qstr("happy", "sad")
# mark one with a plain predicate
mood.add_phase(lambda m: pi if m == "happy" else 0)
# back to a plain Python string
print(mood.observe())
```

## What Rhyme does for you

> [card] Your domain, not bitstrings
>
> The possibilities in a quantum value are your own values. Eight words are eight states, not three qubits, so there is no padding, no encoding tables and no decoding of counts back into names.

> [card] A few operations cover the physics
>
> `add_phase` marks values, `interfere` and `rotate` mix them, and `observe` and `measure` bring you back to classical answers. An oracle is a Python predicate.

> [card] See what your program does
>
> The simulator keeps an explicit map from each value to its amplitude, so you can follow superposition, interference and entanglement in the debugger, step by step.

> [card] Compile and run on real hardware
>
> Functions marked with `@qpu` compile to CUDA-Q or Qiskit and run on simulators or on QPUs from IBM Quantum, Quandela, IonQ, IQM and other providers available through CUDA-Q.

## Where Rhyme is heading

Our vision is a world where every developer can use a quantum computer as easily as a CPU or a GPU. Today about 1,000 people can program a quantum computer, gate by gate. Quantum leaves the lab when the 47 million developers who already build software can call, test and deploy it as just another backend.

Rhyme stays hardware-agnostic on the way there. Several hardware modalities will co-exist, so you write your program once and compile it to the stack and QPU that fit your problem.

**1. From launch to the end of 2026: early access**

Public launch at NVIDIA GTC Berlin, with early access, the reference documentation, advanced CUDA-Q and basic Qiskit compiler prototypes, ten sample algorithms and a VS Code extension prototype.

**2. 2027: first pilots with design partners**

An enterprise compiler and a Qiskit Function, a productivity benchmark of Rhyme against gate-level code, an AI coding agent, a library collection with a hardware recommendation engine, and the first projects with industry design partners, starting in Q2 2027.

**3. 2028: an enterprise platform**

DevOps, testing, CI/CD and governance for quantum programs, the first enterprise deployments, a self-service tier for power users, third-party error correction and compilation optimisation, and libraries ready for quantum-advantage use cases.

> [!NOTE]
>
> These milestones follow the RhymeQ product roadmap of October 2026 and are indicative.

## Where to start

> [card] [Early access](https://docs.rhymeq.com/early-access.md)
>
> What the early-access cohort is, what this release includes and how to share feedback.

> [card] [Getting started](https://docs.rhymeq.com/getting-started.md)
>
> Set up a GitHub Codespace and debug your first Rhyme examples.

> [card] [Tutorials](https://docs.rhymeq.com/basic-tutorials/01_qint.md)
>
> Ten short tutorials, from quantum integers to the Deutsch-Jozsa algorithm and fractals.

> [card] [Language reference](https://docs.rhymeq.com/language-reference/index.md)
>
> The syntax and semantics of Rhyme, type by type and operation by operation.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.