--- title: "torch Breaking Changes on PyPI | Since.dev" description: "What each torch release removed, renamed or changed: 103 breaking changes in 7 versions on PyPI, each with its source." canonical_url: https://since.dev/packages/pypi/torch last_updated: 2026-07-08 --- # torch breaking changes Every breaking API change we have recorded in torch on PyPI: 103 breaking changes in 7 versions. Below are its newest 6 versions, removals and renames first, each with its source. Newest change July 8, 2026 ## 2.13.0 17 breaking changes, first published July 8, 2026. - **Removed** The Bazel build files and CI jobs were removed. Build from source with the CMake/pip flow instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** PyTorch 2.13 no longer ships cp313t wheels. Free-threaded users should move to Python 3.14t. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `c10.StorageImpl.is_cow` Out-of-tree C++ backends calling is_cow must migrate to the new materializer API. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `c10.StorageImpl.maybe_materialize_cow` Out-of-tree C++ backends must register a materializer instead of calling this method. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `Tensor.align_as` Named tensor APIs are gone. Track dimensions by position instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `Tensor.align_to` Named tensor APIs are gone. Track dimensions by position instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `Tensor.names` Named tensor APIs are gone. Track dimensions by position instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `Tensor.refine_names` Named tensor APIs are gone. Track dimensions by position instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `Tensor.rename` Named tensor APIs are gone. Track dimensions by position instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) - **Removed** `torch.align_tensors` Named tensor APIs are gone. Track dimensions by position instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.13.0) And 7 more in this version. ## 2.12.1 1 breaking change, first published June 17, 2026. - **Removed** Release v2.12.1 no longer ships binary builds for CPython 3.13t (free-threaded). Users on that Python version need another install path or a different Python version. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.1) ## 2.12.0 13 breaking changes, first published May 13, 2026. - **Behavior changed** Source builds with SVE enabled are now validated more strictly and may fail during CMake configuration. Fix the compiler/toolchain flags or disable SVE for that build. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** MPS tensors are always allocated in unified memory now. Code relying on device-only placement may see different performance. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** Compiled autograd with FSDP2 needs graph breaks around FSDP2 hooks, or compiled autograd must be disabled for the FSDP2 training step. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** `torch.distributed.nn.functional` Ops in torch.distributed.nn.functional now raise RuntimeError when called inside torch.compile. Migrate to the functional collectives API. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** `torch.linalg.cholesky_inverse` On CUDA this call uses the cuSolver implementation. Code relying on MAGMA-specific numerical behavior or performance may need review. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** `torch.linalg.cholesky_solve` On CUDA this call uses the cuSolver implementation. Code relying on MAGMA-specific numerical behavior or performance may need review. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** `torch.linalg.eigh` On CUDA this call uses the cuSolver implementation. Code relying on MAGMA-specific numerical behavior or performance may need review. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** `torch.linalg.lu_solve` On CUDA this call uses cuSolver/cuBLAS. Code relying on MAGMA-specific numerical behavior or performance may need review. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Behavior changed** `torch.utils.cpp_extension` The cpp_extension and cpp_builder utilities now build extensions as C++20. Extensions that assume C++17 may need updating. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) - **Default changed** `torch._inductor.config.max_autotune_defer_layout_freezing` The 2.11 deferred layout freezing for max_autotune is off by default again. Set max_autotune_defer_layout_freezing to True to keep the deferred behavior. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.12.0) And 3 more in this version. ## 2.11.0 22 breaking changes, first published March 23, 2026. - **Removed** Pre-built CUDA 12.8 and 12.9 binaries no longer support Volta GPUs such as V100. Volta users should install the CUDA 12.6 builds or build from source. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Removed** `torch.ao.quantization.pt2e` torch.ao.quantization.pt2e is gone from PyTorch. Install torchao and import the same APIs from torchao.quantization.pt2e. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Removed** `torch.ao.quantization.quantizer` The quantizer module is gone from PyTorch. Install torchao and import from torchao instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Removed** `torch.export.export_for_training` Use torch.export.export instead, which returns the same graph as the previous export_for_training. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Removed** `torch.nn.attention.varlen_attn` Causal attention is now expressed with window_size=(-1, 0) instead of is_causal=True. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Removed** `torch.onnx.export` Remove fallback=True or fallback=False from calls. Implement any fallback to the legacy exporter in your own code. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Behavior changed** Plain pip install torch now pulls CUDA 13.0 wheels, which only support Turing and newer GPUs on Linux x86_64. Users with older drivers or GPUs must pass a CUDA 12.x index URL. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Behavior changed** `_PhiloxState.offset` Code that reads offset as an int must call .item(). Setting it requires a tensor. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Behavior changed** `_PhiloxState.seed` Code that reads seed as an int must call .item(). Setting it requires a tensor. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) - **Behavior changed** `DebugInfoWriter` Debug info is now written under $XDG_CACHE_HOME/torch when that variable is set, instead of always under \~/.cache/torch. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.11.0) And 12 more in this version. ## 2.10.0 10 breaking changes, first published January 21, 2026. - **Removed** `nn.attention.flex_attention` Nested Jagged Tensor support was removed from nn.attention.flex_attention. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Removed** `torch.utils.data.datapipes.iter.grouping` The imports SHARDING_PRIORITIES and ShardingFilterIterDataPipe from torch.utils.data.datapipes.iter.grouping are no longer supported. Import them from torch.utils.data.datapipes.iter.sharding instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Renamed** `pytorch-triton` The package pytorch-triton was renamed to triton. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Behavior changed** `DeviceMesh._flatten` Implicit slicing of a flattened dim from the root mesh now emits a warning and will eventually be removed; keep your own bookkeeping of the flattened mesh. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Behavior changed** `Sampler` Custom samplers that use the data_source argument must be updated accordingly. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Behavior changed** `TORCH_CHECK_{COND}` TORCH_CHECK\_{COND} behavior changed to be non-fatal, so error handling in C++ code may change. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Behavior changed** `torch.compile` The aot_eager backend may now give results that differ from before because extra compile-only decompositions no longer run. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Behavior changed** `torch.nn.functional.flex_attention` Using a sliced BlockMask in nn.functional.flex_attention now raises an error. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Default changed** `torch._inductor.config.combo_kernels` Turn on LOAF for OSS by default; behavior of compiled code may change. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) - **Default changed** `torch.onnx.export` torch.onnx.export uses fallback=False by default; set fallback=True manually to keep 2.9 behavior. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.10.0) ## 2.9.0 18 breaking changes, first published October 15, 2025. - **Removed** MPS is only supported on macOS 14 or later, so older macOS versions such as Ventura can no longer use it with this release. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** The minimum Python version for PyTorch 2.9.0 is 3.10. Users on older Python must upgrade before installing. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.compile` The experimental ONNX Runtime compile backend is no longer supported. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.export.draft_export` The draft export fallback is no longer triggered automatically; set TORCH_ONNX_ENABLE_DRAFT_EXPORT=True to opt in. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.export.export_for_training` Use torch.export.export instead; export_for_training is deleted. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.onnx.dynamo_export` Use torch.onnx.export instead. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.onnx.enable_fake_mode` The function is removed; dynamo=True exports use FakeTensors by default. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.onnx.symbolic_caffe2` Caffe2 support in the ONNX exporter has ended and the module is removed. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.onnx.symbolic_opsets` Private torch.onnx.symbolic_opsets functions can no longer be accessed; copy the source if needed. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) - **Removed** `torch.onnx.verification` Deprecated members in torch.onnx.verification are removed. [github.com](https://github.com/pytorch/pytorch/releases/tag/v2.9.0) And 8 more in this version. ## Check your code These are the changes. To see which of them reach your code, [scan a public repository](https://since.dev/scan.md), or have your coding agent call `check_code` with your manifest or lockfile over [MCP](https://since.dev/docs.md#mcp). [The API](https://since.dev/docs.md#api) reads every change in every version, not only the newest. On Team and up, a fix pull request opens for the changes that reach a connected repository. Nothing merges or deploys. ## Put your code on continuous management. Scan a public repository now, or give your agent one line. ### See how out of date your code is. [Scan a repository](https://since.dev/scan.md). Any public GitHub repository. No sign-in. ### Tell your agent what changed since it was trained. `claude mcp add --transport http since https://api.since.dev/mcp` --- Every page on this site as Markdown: https://since.dev/sitemap.md. For agents: https://since.dev/llms.txt.