qiskit
Qiskit is an open-source SDK for working with quantum computers at the level of extended quantum circuits, operators, and primitives.
- 8k
- Python
- Apache-2.0
- 6 h ago
The documentation content home for https://quantum.cloud.ibm.com/docs
Repo Qiskit/documentation, read from GitHub 1 min ago. Qiskit has not joined Take My Repo.
Qiskit/documentation is a public repo on GitHub by Qiskit. The documentation content home for https://quantum.cloud.ibm.com/docs. It declares the Apache-2.0 license on GitHub. It is written mainly in Jupyter Notebook, has 112 stars and last changed on 2026-11-20. Nobody has reviewed it on Take My Repo yet, which says nothing about its quality. Read from GitHub on 2026-11-22.
In the maintainer's words. Shown from GitHub; Take My Repo does not keep a copy.
The documentation content home for https://quantum.cloud.ibm.com/docs and https://docs.quantum.ibm.com (excluding API reference).
Preview the main branch at the staging link: https://qiskit.github.io/documentation/main
Refer to:
This repository is dual-licensed to distinguish between code and content.
LICENSE file for more informationLICENSE-DOCS file for more information.Maintaining up-to-date documentation is a huge challenge for any software project, but especially for a field like quantum computing, because advances in new research and technological capabilities come at a fast pace. As a result, we greatly appreciate anyone who takes the time to support us in keeping this content accurate and up to the highest quality standard possible, to benefit the broadest range of users.
Read on for more information about how to support this project:
This is the quickest, easiest, and most helpful way to contribute to this project and improve the quality of Qiskit® and IBM Quantum® documentation. There are a few different ways to report issues, depending on where it was found:
If you think there are gaps in our documentation, or sections that could be expanded upon, we invite you to open a new content request issue.
Not every new content suggestion is a good fit for docs, nor are we able to prioritize every request immediately. However, we will do our best to respond to content requests in a timely manner, and we greatly appreciate our community's efforts in generating new ideas.
If you are interested in writing the new content yourself, or already have some draft work you think could be integrated, please also mention that in the issue description. If your content suggestion is accepted, we will let you know and work with you to get the content written and reviewed.
Please note: we DO NOT accept unsolicited PRs for new pages or large updates to existing content. The content that we include in docs is carefully planned and curated by our content team and must go through the appropriate review process to ensure the quality is of the highest possible standard before deploying to production. As a result we are very selective with which content suggestions are approved, and it is unlikely that PRs submitted without an associated approved content request will be accepted.
You can help the team prioritize already-open issues by doing the following:
See our Contribution policy for information on contributing fixes.
This repository is very large, so it is slow to clone the first time. We recommend instead using the argument --filter=blob:none:
git clone --filter=blob:none https://github.com/Qiskit/documentation.git
--filter=blob:none means that Git will lazily download file contents when you need them, rather than eagerly downloading everything on the initial clone.
We also recommend running this command once to tell Git to ignore the gh-pages branch, which is solely used for PR previews and is very large:
git config --add remote.origin.fetch '^refs/heads/gh-pages'To contribute code to this repository, your email must be configured properly with Git, as follows:
git config --get user.email. You do not need to make any further changes if the email address is one of the addresses set up with your github.com account.git config --global user.email <your-email>. Use an email address associated with your github.com account.We provide several tools to preview the documentation locally and to run quality checks. While many of these tools run in your PR through CI, we recommend installing them locally for faster iteration.
First, install the below software:
pipx install tox.Then, install the dependencies with:
npm install
You can preview the docs locally by following these two steps:
./start in your terminal, and open http://localhost:3000 in your browser.
python start instead. Alternatively, use Windows Subsystem for Linux and run ./start.The preview application does not include the top nav bar. Instead, navigate to the folder you want with the links in the home page. You can return to the home page at any time by clicking "IBM Quantum Documentation Preview" in the top-left of the header.
Maintainers: when you release a new version of the image, you need to update the image digest in ./start by following the instructions at the top of the file and opening a pull request.
Occasionally, Docker might fail when it runs out of disk space. For example, you might encounter an error like this:
ENOSPC: no space left on device, mkdir '/tmp/xfs-4d1dfe51'
Try running docker system prune to clear Docker's system space.
API docs authors can preview their changes to one of the APIs by using the -a parameter to specify the path to the docs folder:
npm run gen-api -- -p <pkg-name> -v <version> -a <path/to/docs/_build/html>../start and open up http://localhost:3000, as explained in the prior section.Addon docs authors can preview guide and how-to changes by pointing the addon pipeline at a local Sphinx build:
npm run gen-addon -- -p <pkg-name> --sphinx-artifact-folder <path/to/docs/_build/html>../start and navigate to http://localhost:3000/docs/addons/<pkg-name>.We use multiple tools to ensure that documentation meets high standards. These tools will run automatically in your PR through CI, but it is much faster to run the checks locally.
Use ./fix to automatically apply fixes, like fixing formatting. Warning: we cannot automatically fix every problem, so you should also run ./check afterwards.
./fix
Use ./check to validate that there are no issues. If you encounter an error, fix it by following the instructions in the error message, then keep running ./check and fixing any errors until it fully passes without any error.
./check
On Windows, run python fix and python check instead. Alternatively, use Windows Subsystem for Linux and run ./fix and ./check.
You may find it convenient to install the following VSCode extensions to automatically run some of our tools. Setting up these extensions is optional.
We offer some tools that are not included in ./check and ./fix. Likewise, many of the checks skip API docs by default.
Run npm run to see a list of all our checks. For any particular check, run npm run my-check -- --help for more information and advanced arguments, such as npm run check:markdown -- --help.
When adding a new notebook, you'll need to tell the testing tools how to handle it.
To do this, add the file path to scripts/config/notebook-testing.toml. There are
four categories:
[groups.normal]: Notebooks to be run normally in CI. These notebooks
can't submit jobs as the queue times are too long and it will waste
resources. You can interact with IBM Quantum to retrieve jobs and backend
information.[groups.local-sim]: Notebooks that submit jobs, but that are small
enough to run on a 5-qubit simulator. We will test these notebooks in CI by
patching least_busy to return a 5-qubit fake backend.[groups.test-eagle]: These notebooks can't run with a local simulator, but
can be mocked with the test-eagle device, which returns nonsense results.
We can trigger a manual job to test these with test-eagle.[groups.cron-job-only]: For notebooks that can't be tested using the 5-qubit
simulator patch. We skip testing these in CI and instead run them twice per
month. Any notebooks with cells that take more than five minutes to run are
also deemed too big for CI. Try to avoid adding notebooks to this category if
possible.[groups.exclude]: We never test these notebooks.If you don't do this step, you will get the error "FAILED scripts/nb-tester/test/test_notebook_classification.py::test_all_notebooks_are_classified".
If your notebook requires extra dependencies (such as latex or graphviz), you
must also add it to the EXTRA_DEPS_NOTEBOOKS list in
.github/workflows/main.yml. We don't install these for every job because they
take a while to install and only a handful of notebooks use them. You will know
when you need them because CI will fail.
Add a new markdown cell under your title with a version-info tag. When you
execute the notebook (see the next section), the script will populate this cell
with the package versions so users can reproduce the results. Do not add
anything else to this cell as the contents of the cell will be completely
overwritten when it's next executed.
The package versions will also update during the cron jobs.
Before submitting a new notebook or code changes to a notebook, you must run
the notebook using tox -- --write <path-to-notebook> and commit the results.
If the notebook submits jobs, also use the argument --test-strategy=hardware.
This means we can be sure all notebooks work and that users will see the same
results when they run using the environment we recommend.
We use tox to execute notebooks in a reproducible Python environment. First,
install tox using pipx:
pipx install tox
You may also need to install a few system dependencies: TeX, Poppler, and
graphviz. On macOS, you can run brew install mactex-no-gui poppler graphviz.
On Ubuntu, you can run apt-get install texlive-pictures texlive-latex-extra poppler-utils graphviz.
tox
tox -- <path/to/notebook.ipynb> <path/to/another-notebook.ipynb>
--write argument. (Note this generates the package version info cell mentioned in this section.)
Since we only allow writing results from real hardware, you will usually also
need to pass --test-strategy=hardware. Note this means the run will use QPU
time.
tox -- optional/paths/to/notebooks.ipynb --write --test-strategy=hardware
When you make a pull request changing a notebook that doesn't submit jobs, you can get a version of that notebook that was executed by tox from CI. To do this, click "Show all checks" in the info box at the bottom of the pull request page on GitHub, then choose "Details" for the "Test notebooks" job. From the job page, click "Summary", then download "Executed notebooks". Otherwise, if your notebook does submit jobs, you need to run it locally using the steps mentioned earlier.
We don't want users to see warnings that can be avoided, so it's best to fix
the code to avoid them. However, if a warning is unavoidable, you can stop it
blocking CI by adding an ignore-warnings tag to the cell. In VSCode,
right-click the cell, choose "Add cell tag", type ignore-warnings, then press
"Enter". In Jupyter notebook (depending on version), choose View > Right
Sidebar > Show Notebook Tools, then under "Common Tools" add a tag with text
ignore-warnings.
Our CI checks notebooks run from start to finish without errors or warnings. You can add extra checks in notebooks to catch other unexpected behavior.
For example, say we claim a cell always returns the string 0011. It would be
embarassing if this was not true. We can assert this in CI by adding the
following code cell, and hide it from users with a remove-cell tag.
# Confirm output is what we expect. assert _ == '0011'
In Jupyter notebooks, the underscore _ variable stores the value of the
previous cell output. You should also add a comment like
# Confirm output is what we expect so that authors know this
block is only for testing. Make sure you add the tag remove-cell.
If something ever causes this value to
change, CI will alert us.
Run npm run fmt to automatically format the README, .github folder, and scripts/ folder. You should run this command if you get the error in CI run Prettier to fix.
To check that formatting is valid without actually making changes, run npm run check:fmt.
This is useful when we make improvements to the API generation script.
You can regenerate all API docs versions following these steps:
main using git checkout -b <branch-name>.git status and creating a new commit for them if necessary.npm run regen-apis to regenerate all API docs versions for qiskit, qiskit-ibm-runtime, and qiskit-ibm-transpiler.Each regenerated version will be saved as a distinct commit. If the changes are too large for one single PR, consider splitting it up into multiple PRs by using git cherry-pick or git rebase -i so each PR only has the commits it wants to target.
If you only want to regenerate the latest stable minor release of each package, then add --current-apis-only as an argument, and in case you only want to regenerate versions of one package, then you can use the -p <pkg-name> argument.
Alternatively, you can also regenerate one specific version:
qiskit, qiskit-ibm-runtime, or qiskit-ibm-transpiler) and its version.npm run gen-api -- -p <pkg-name> -v <version>,
e.g., npm run gen-api -- -p qiskit -v 0.45.0If the version is not for the latest stable minor release series, then add --historical to the arguments. For example, use --historical if the latest stable release is 0.45.* but you're generating docs for the patch release 0.44.3.
Additionally, If you are regenerating a dev version, then you can add --dev as an argument and the documentation will be built at /docs/api/<pkg-name>/dev. For dev versions, end the --version in -dev, e.g. -v 1.0.0-dev. If a release candidate has already been released, use -v 1.0.0rc1, for example.
In this case, no commit will be automatically created.
Addon docs for packages like are generated by a separate pipeline from the API pipeline. Content is published to docs/addons/{pkg}/ rather than docs/api/{pkg}/.
Run:
npm run gen-addon -- -p <pkg-name>
To regenerate all addon packages at once:
npm run gen-addon -- --all
Or if you don't need to download but need to re-run the pipeline you can pass "--skip-download" to speed things up
npm run gen-addon -- --all --skip-download
Open a pull request with the generated changes.
0.5.zip, and upload the renamed zip file to https://ibm.ent.box.com/folder/246867452622.
scripts/config/api-html-artifacts.json (see how to get the shareable link here). Copy an existing section, like qiskit-addon-paulice, then update it with the correct information. Save a commit with the message "Set up Box link".Pkg.ADDON_NAMES in scripts/js/lib/api/Pkg.ts and add a new Pkg(...) with the correct title, githubSlug, and language in that file as well. Save a commit with the message "Add package name". Note: The package name becomes the URL. Contact the Addon's author for the correct package name.gen-api (as described below), then gen-addon (See above section)) Note that these pipeline commands require major.minor.update for the file name. For example, npm run gen-addon -- -p qiskit-fermions -v 0.1.0. This is different from the zip file's name. Check in the changes and list the commands you ran in the PR description.npm run check:internal-links -- --current-apis, npm run check:spelling, npm run check:markdown -- --apis, and ./check. Fix any issues directly in the generated content. Note any problems with link generation etc. and let the frontend team know, in case they need to fix the pipeline../start --apis (Docker must be running through Rancher or similar.) Navigate to docs/addons/{pkg}/ in the preview to see the new content.Use this process when we want to publish new API docs, such as when we release a new version of a package like Qiskit SDK.
You must have the environment variable GITHUB_TOKEN saved to your environment. You can check if it is by running echo $GITHUB_TOKEN in your terminal.
The easiest way to set up a new token is with the GitHub CLI:
gh auth login. Follow the instructions to log-in in to github.com (not enterprise GitHub).gh auth status --show-token. Find the token for github.com; make sure it is not for enterprise GitHub.GITHUB_TOKEN to your environment, such as by updating ~/.zprofile with export GITHUB_TOKEN=<token>.The process depends on which type of release you are generating. Some key terms to know:
qiskit/documentation repository, the latest docs are at the top-level of the API folder, e.g., the files at docs/api/qiskit (not docs/api/qiskit/dev or docs/api/qiskit/2.3).qiskit/documentation repository, historical docs are stored in a folder with their number, such as docs/api/qiskit/0.46 and docs/api/qiskit/2.3.main branch several times a week; the other packages do not have dev docs. In the qiskit/documentation repository, dev docs are stored in the dev folder, e.g., docs/api/qiskit/dev.dev/ folder, like docs/api/qiskit/dev.All release types start with the following steps:
In Git, check out the main branch of Qiskit/documentation and pull any updates. Then, create a new Git branch.
Determine which documentation you want to generate (e.g., qiskit or qiskit-ibm-runtime) and its full version, e.g., 0.45.2 or 1.2.0rc1.
Download a CI artifact with the project's documentation. To find this:
Find the relevant GitHub Actions workflow for the project:
Find the run for the release by looking at the middle column with the blue text; look for the version number, like 0.45.2. For the rc1 release, look for main in blue text and a run description like "Prepare 2.3.0rc1".
Click the CI run name. (Not the middle column with the blue link!)
In the left navbar, it should show as selected the "Summary" page with the house.
Scroll down to "Artifacts" and look for the artifact related to documentation, such as html_docs.
Download the artifact by clicking on its name. Note: For qiskit-addon-utils (and possibly others) the downloaded zip file might contain an extra parent directory. This will cause the gen-addon script to run incorrectly and leave your changes in a broken state. To fix this, uncompress the zip file and open the top-level directory, so you are viewing a collection of files and folders such as _downloads, _images, objects.inv, and others. Select all these files and folders and compress into a new zip file, which you will rename (see step 5 below) and upload (step 6).
On some operating systems, the downloaded zip file will be auto-expanded rather than staying a zip file. If this happens, compress it back to a zip file. On macOS, secondary-click on the folder in Finder and use the "Compress" option.
Rename the downloaded zip file with its minor-version number. For example, for the release 0.45.2, rename html_docs.zip to 0.45.zip. For release candidates (rc), use a value like 2.3-rc.zip.
Upload the renamed zip file to https://ibm.ent.box.com/folder/246867452622. If this is a patch release, this step will overwrite the prior file; if it's the rc1 release or a new minor version like 2.3.0, it will be a new file.
Proceed with the final steps for the type of release (patch, rc1, or major/minor; see the relevant section below).
Examples of when to use this process:
2.3.0, and 2.3.1 is released2.3.2 to 2.3.32.3.1rc1, and 2.3.1.rc2 is releasedSteps:
npm run gen-api -- -p <pkg-name> -v <version> --dev, e.g., npm run gen-api -- -p qiskit -v 2.3.0rc2 --dev.npm run gen-api -- -p <pkg-name> -v <version> --historical, e.g., npm run gen-api -- -p qiskit -v 2.3.2 --historical.npm run gen-api -- -p <pkg-name> -v <version>, e.g., npm run gen-api -- -p qiskit -v 2.3.2.-p qiskit-c instead of -p qiskit.This process is only for the first release candidate (rc1). Subsequent release candidates like rc2 should use the process for patch releases.
Copy shared link button.People with the link from the menu under "Share Link" (default is Invited people only) and go to Link Settings.Link Expiration, select Disable Shared Link on and set an expiration date of ~10 years into the future. (There must be an expiration date.)Shared Link Settings tab. Do not use the link from the prior screen.scripts/config/api-html-artifacts.json and find the dev entry for the package, like qiskit.dev link with the Box link, rather than the GitHub link.npm run gen-api -- -p <pkg-name> -v <version> --dev, e.g., npm run gen-api -- -p qiskit -v 2.3.0rc1 --dev.-p qiskit-c instead of -p qiskit.Examples of when to use this process:
2.3.1, and 2.4.0 is released2.3.1, and 3.0.0 is releasedCopy shared link button.People with the link from the menu under "Share Link" (default is Invited people only) and go to Link Settings.Link Expiration, select Disable Shared Link on and set an expiration date of ~10 years into the future. (There must be an expiration date.)Shared Link Settings tab. Do not use the link from the prior screen.scripts/config/api-html-artifacts.json by adding the new version with the direct link from step 9.scripts/js/lib/api/Pkg.ts file to add name and entry there (see the instructions in Step 4 of Add a new addon package)._package.json file in the package's top-level folder, such as docs/api/qiskit/_package.json.npm run gen-api -- -p <pkg-name> -v <version> --historical, using the version from the previous step. For example, npm run gen-api -- -p qiskit -v 0.2.1 --historical.-p qiskit-c instead of -p qiskit.npm run gen-api -- -p <pkg-name> -v <version>, using the version from the new release. For example, npm run gen-api -- -p qiskit -v 0.3.0.-p qiskit-c instead of -p qiskit.main branch by looking at the middle column with the blue text; look for main.html_docs.6026447195 from the link https://github.com/Qiskit/qiskit/actions/runs/23345366690/artifacts/6026447195.api-html-artifacts.json, update the dev entry with the following value. Replace <NUMBER> with the number from the prior step. Qiskit: https://api.github.com/repos/Qiskit/qiskit/actions/artifacts/<NUMBER>/zip. Qiskit Runtime: https://api.github.com/repos/Qiskit/qiskit-ibm-runtime/actions/artifacts/<NUMBER>/zip-dev. For example, if the latest release is 2.3.0, then the dev version would be 2.4.0-dev.npm run gen-api -- -p <pkg-name> -v <version> --dev, e.g., npm run gen-api -- -p qiskit -v 2.4.0-dev --dev.-p qiskit-c instead of -p qiskit./dev folder.objects.inv (advanced)Since objects.inv is compressed, we can't review changes through git diff. Git does tell you if the file has changed, but this isn't that helpful as the compressed file can be different even if the uncompressed contents are the same.
If you want to see the diff for the uncompressed contents, first install sphobjinv.
pipx install sphobjinv
The add the following to your .gitconfig (usually found at ~/.gitconfig).
[diff "objects_inv"]
textconv = sh -c 'sphobjinv convert plain "$0" -'
When a new version of an API is released, we should also update nb-tester/requirements.txt to ensure that our notebooks still work with the latest version of the API. You can do this upgrade either manually or wait for Dependabot's automated PR.
CI will fail on Dependabot's PR at first due to not having access to the token. To fix this, have someone with write access close and then immediately reopen the PR to trigger CI.
You can land the API generation separately from the requirements.txt version upgrade. It's high priority to get out new versions of the API docs ASAP, so you should not block that on the notebook version upgrade if you run into any complications like failing notebooks.
See the section "Syncing content with open source repo" in the internal docs repo's README.
Qiskit/documentation is a public repo on GitHub by Qiskit. The documentation content home for https://quantum.cloud.ibm.com/docs.
It declares the Apache-2.0 license on GitHub.
Its maker tagged it for macOS, Windows and Linux, or as Docker or cross-platform. That is declared by the maker; Take My Repo has not tested it.
Open "Take it" on its page: Take My Repo writes an install plan for your machine and hands it to your AI agent (Claude Code, Codex or Cursor), which asks you before each step. Take My Repo has not tested it on any machine.
Its last change on GitHub was on 2026-11-20. GitHub counts 213 open issues and pull requests. Take My Repo does not judge from that whether it is maintained.
Nobody has reviewed it on Take My Repo yet, which says nothing about its quality.
Qiskit is an open-source SDK for working with quantum computers at the level of extended quantum circuits, operators, and primitives.
Qiskit is an open-source SDK for working with quantum computers at the level of circuits, algorithms, and application modules.
A collection of Jupyter notebooks showing how to use the Qiskit SDK
A high performance Python graph library implemented in Rust.
Aer is a high performance simulator for quantum circuits that includes noise models
This page is also available as structured data, with the same sources and dates.
/api/v1/resources/gh/Qiskit/documentationTake My Repo has not tested this on your machine. Requirements come from the project's own documentation.