NewYour coding agent can read the release notes before it upgrades.Set up the MCP server โ
PyPI, #773 most downloaded on PyPI
A refreshing functional take on deep learning, compatible with your favorite libraries
Latest stable
10.0.0
Last release 26 days ago
27 Oct 2026
Release timing varies
gaps range from 1 weeks to 7 months
Nearly every release is documented
notes for 58 of the last 60 stable releases
3 versions withdrawn
withdrawn after publishing
12 years old
252 releases ยท first in 2014
Drag across the ruler to compare, or tap a release to open it.
Fix backprop_reduce_max GPU kernel.
backprop_reduce_max GPU kernel.Update to support torch v1.9.0.
Fix issue #496: Avoid Pydantic versions with security vulnerabilities.
tuplify layer.concatenate layer.resizable layer.force parameter for model.set_dim().with_getitem.@adrianeboyd, @honnibal, @ines, @kludex, @polm, @svlandeg, @thomashacker
Your coding agent can read these notes before it upgrades. Set up the MCP server โ
Fix issue #486: Fix expand_window for empty docs on GPU
expand_window for empty docs on GPUcatalogue>=2.0.3 due to performance regressions related to importlib-metadataFix issue #465: Fix saving models to Pathy paths
map_list layer (#472)require_cpuFix issue #464: Fix list2padded op
nO to LayerNorm> ๐ฎ This version of Thinc has been rewritten from the ground up and will be used to power the upcoming spaCy v3.0. The new Thinc v8.0 is a lightweight
๐ฎ This version of Thinc has been rewritten from the ground up and will be used to power the upcoming spaCy v3.0. The new Thinc v8.0 is a lightweight deep learning library that offers an elegant, type-checked, functional-programming API for composing models, with support for layers defined in other frameworks such as PyTorch, TensorFlow or MXNet. You can use Thinc as an interface layer, a standalone toolkit or a flexible way to develop new models. For more details, see the documentation.
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
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
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
Updates for Python 3.10 and 3.11 (#791):
wrapt to v1.14.1.@adrianeboyd, @honnibal, @ines
Fix numpy compatibility in binary wheel releases.
numpy compatibility in binary wheel releases.cupy-cuda111 extra requirement.Update for compatibility with cupy v8.
cupy v8.PyTorchWrapper.numpy build constraints from pyproject.toml.Add seed argument to ParametricAttention.
seed argument to ParametricAttention.numpy headers and add numpy build constraints.hypothesis v5.Beam.Restrict compatible cupy versions to <8.0.0.
cupy versions to <8.0.0.One column per quarter.