FlagQuantum v0.3.1

FlagQuantum 0.3.1 adds the approved local research-capability services that landed after the 0.3.0 distributions were built.

• Discover and describe bounded research capabilities through flagquantum.services.
• Validate and estimate typed requests, execute supported local CPU operations, and validate structured evidence.
• Includes the supplied two-qubit Pauli-sum ground-state VQE path and a versioned evidence bundle.
• Preserves the 0.3.0 public API and Python 3.10–3.12 support.

This release does not construct molecular Hamiltonians from chemical specifications, run on quantum hardware, or include an MCP server or agent runtime. The MCP adapter is a separate package and release.

Install with python -m pip install flagquantum==0.3.1.

FlagQuantum v0.3.0

FlagQuantum 0.3.0 brings the 0.3 release line to a stable package.

• Stable ABI wheels for Python 3.10–3.12 on Linux x86_64, macOS arm64, and Windows amd64.
• PyTorch-first quantum simulation, training, and remote execution.
• Recoverable Jiuding and Quafu jobs with opt-in receipts, plus ordered Quafu multi-circuit results.
• English and Chinese editions of the six-lesson FlagOS 2026 workshop.

Install with python -m pip install flagquantum==0.3.0.

FlagQuantum 0.3.0rc3

FlagQuantum 0.3.0rc3 is the cross-platform release candidate for the 0.3 line.

Highlights:

• one cp310-abi3 wheel per supported platform for Python 3.10-3.12
• verified Linux x86_64, macOS arm64, and Windows amd64 wheel pipelines
• native CPU extension built against the PyTorch 2.10 Stable ABI baseline
• runtime support validated on PyTorch 2.13 and 2.14
• MSVC portability fixes for Stable ABI flags, OpenMP SIMD, and bit lookup

The existing v0.3.0rc2 release remains immutable.

Install after publication:

python -m pip install --pre flagquantum==0.3.0rc3

FlagQuantum 0.3.0rc2

FlagQuantum 0.3.0rc2 is the corrected 0.3 release candidate.

Highlights:

• one cp310-abi3 wheel per supported platform for Python 3.10-3.12
• native CPU extension built against the PyTorch 2.10 Stable ABI baseline
• runtime support validated on PyTorch 2.13 and 2.14
• self-contained flagquantum-compiler-qsteed 0.2.0 integration for development images
• direct Jiuding and Quafu workflow improvements included from the 0.3 release line

The existing v0.3.0rc1 prerelease is preserved as an immutable incomplete release record. No 0.3.0rc1 distribution was uploaded to PyPI.

Install after publication:
python -m pip install --pre flagquantum==0.3.0rc2

FlagQuantum 0.3.0rc1

FlagQuantum 0.3.0rc1 is the first release candidate for the 0.3 series. It packages the user-visible work completed since 0.2.0 for installation and provider-workflow validation before the final 0.3.0 release.

Highlights include expanded circuit composition and output APIs, gradient and vector-derivative APIs, OpenQASM import, compiler optimization levels, distributed workflows, digital-twin validation, and updated Jiuding and Quafu provider guidance.

Install the release candidate with:

python -m pip install --pre flagquantum==0.3.0rc1

Known issue: flagquantum-compiler-qsteed 0.1.1 declares compatibility with FlagQuantum 0.2.x only, so QSteed development images do not yet resolve with this release candidate. Do not bypass dependency checks; users who require that external adapter should remain on FlagQuantum 0.2.0 until a compatible adapter release is available.

See docs/reference/RELEASE_NOTES.md in this tag for the complete change list.

FlagOS 2.1 — FlagQuantum v0.1.0

Release v0.1.0

Initial Release

Other

• Add the initial impl (#1) by @FlagQuantum
• Feat: Add quantum circuit visualization and 6 tutorial examples (#2) by @FlagQuantum
• Feat: Add OpenQASM 3.0/2.0 exporter for real quantum hardware and cross-framework execution (#3) by @FlagQuantum
• Fix: add Mthreads GPU support and improve device detection in v0.1.0-rc0 (#11) by @FlagQuantum
• Docs: add multi-chip backend support info to features list (#12) by @FlagQuantum

Contributors

Thanks to all contributors who made this release possible:

• @FlagQuantum (5 PRs)

---

Generated by Release Notes Generator