> For the complete documentation index, see [llms.txt](https://docs.quantaria.xyz/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.quantaria.xyz/roadmap.md).

# Roadmap

Prove it small, then scale it

Quantaria's plan is to prove the approach at small scale, then scale it. Each step publishes its results, including the ones that do not work. The roadmap is ordered by stage rather than by date, and it may change as research results come in.

| Stage    | Work                                                                                                                                                                                             | Status   |
| -------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -------- |
| Now      | **Hilbert-1 Nano and Base.** Live on the [Quantaria Console](/quantaria-console.md). Benchmarks against dense transformers of the same size follow: loss curves, perplexity, FID and memory use. | Live     |
| Next     | **Bond, Schmidt and Hilbert-1 Large.** Open-source Bond. Train Schmidt for on-device long context. Scale Hilbert to Large, with the scale set after Base results.                                | Planned  |
| Then     | **Dirac hybrid layers.** Swap selected attention layers for disentangler circuits on simulators, then on cloud quantum hardware.                                                                 | Research |
| Research | **Quantum-native training.** Area-law rank allocation from measured entanglement entropy, DMRG-style sweeps alongside gradient descent, and Born-machine heads.                                  | Research |

## What each stage must show

A stage is complete when its results are published, not when the model exists. The exit criteria below are the public commitments for each stage.

* **Now (Hilbert-1 Nano and Base).** Publish loss curves, perplexity, FID and memory use against a dense transformer of the same size, trained on the same data. Include results that do not favour the tensor-network model.
* **Next (Bond, Schmidt and Hilbert-1 Large).** Release Bond with its source code and license. Publish Schmidt's on-device results for long context. Set Large's scale from the Base results and publish the comparison.
* **Then (Dirac hybrid layers).** Show each hybrid layer on circuit simulators first, with the simulator setup stated. Run the layers on cloud quantum hardware only after simulator results are published, and report the effect on quality, including when it is negative.
* **Research (quantum-native training).** Publish methods and experiments before making claims about them. Area-law rank allocation, DMRG-style sweeps and Born-machine heads are open research questions and may not reach production.

## Commitments

* **Matched baselines.** Every model is compared with a dense transformer of the same size, trained on the same data.
* **Cited numbers.** Every research figure links to its paper. See [Research Foundations](https://github.com/bhavy2908/dataai_docs/tree/main/architecture/research-foundations.md).
* **Hardware-honest.** Quantaria says plainly when a result came from a simulator.

Updates are announced at [quantaria.xyz](https://quantaria.xyz). Questions: <hello@quantaria.xyz>.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.quantaria.xyz/roadmap.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.