NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI
PyPI, #2180 most downloaded on PyPI
Efficiently computes derivatives of NumPy code.
Latest stable
1.9.1
Last release 4 months ago
02 Jul 2026
Release timing varies
gaps range from 5 weeks to 2.7 years
Rarely documented
notes for 5 of 33 stable releases
Nothing withdrawn
no release was ever pulled
12 years old
33 releases · first in 2015
Drag across the ruler to compare, or tap a release to open it.
This is a small patch release that fixes a gradient regression in reduction operations.
29 June 2026
This is a small patch release that fixes a gradient regression in reduction operations.
**kwargs, so no longer fail with a TypeError when NumPy passes an out=None argument through to them. This restores differentiation of reductions over ArrayBox values in a few edge cases that broke after #717 (#789, by @fjosw).Full CHANGELOG: https://github.com/HIPS/autograd/compare/v1.9.0...v1.9.1
_23 June 2026, released on PyPI on 26 June 2026_
23 June 2026, released on PyPI on 26 June 2026
We are happy to announce the Autograd v1.9.0 release. It brings broader NumPy 2.x and Python compatibility, several gradient correctness fixes, and better integration with array-like containers such as xarray's DataArray.
This release supports both NumPy 1.x and 2.x. Python versions 3.10 through 3.14 are supported. We have dropped support for Python 3.9, which is now end of life (#748).
xarray.DataArrays that hold ArrayBoxes (#717, #716).'same' mode for scipy.signal.convolve, including its forward pass and VJP (#721).np.where now preserves native Python types instead of promoting them to float64 (#779, fixes #671).np.mean, np.std, np.var, and linalg.norm failing when passed a list or tuple containing ArrayBox objects (#757).np.astype, which was introduced in NumPy 2.0 (#714).np.conjugate (#713).np.absolute gradients so they handle the division-by-zero case the same way np.abs already did (#712).np.empty_like and np.full_like as non-differentiable functions so they pass through tracing cleanly (#710).linalg.norm VJP and JVP for complex numbers (#705, fixes #666).exclude-newer = "3 days ago" to guard against supply-chain attacks during dependency resolution (#775).macos-13 CI image with macos-15-intel (#747).xarray.DataArrays (#782).Full CHANGELOG: https://github.com/HIPS/autograd/compare/v1.8.0...v1.9.0
Your coding agent can read these notes before it upgrades. Set up the MCP server →
Fix numpy.core.einsumfunc deprecation by @fjosw in https://github.com/HIPS/autograd/pull/628
We are happy to announce the release of Autograd version 1.8.0.
This version supports Python versions 3.9–3.13 and includes minor features, some bug fixes, and many maintenance updates.
numpy.core.einsumfunc deprecation by @fjosw in https://github.com/HIPS/autograd/pull/628tanh(x) derivative README example by @agriyakhetarpal in https://github.com/HIPS/autograd/pull/635, https://github.com/HIPS/autograd/pull/636__array_namespace__ to ArrayBox by @yaugenst-flex in https://github.com/HIPS/autograd/pull/647scipy.special.logsumexp type promotion, test against NumPy and SciPy nightlies by @agriyakhetarpal in https://github.com/HIPS/autograd/pull/643scipy.misc by @fjosw in https://github.com/HIPS/autograd/pull/669finally statement by except BaseException to suppress all exceptions by @fjosw in https://github.com/HIPS/autograd/pull/653test_dtypes for now by @agriyakhetarpal in https://github.com/HIPS/autograd/pull/690pypy-3.9 by @agriyakhetarpal in https://github.com/HIPS/autograd/pull/697Full Changelog: https://github.com/HIPS/autograd/compare/v1.7.0...v1.8.0
This release includes support for NumPy version 2.0.0
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
bump version number for new core
bump version number for new core
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
One column per quarter.