vllm changelog & breaking changes

v0.23.0

2026-06-15

⚠ Potential breaking changes

Unified parser**: Reasoning and tool-call parsing are now unified behind a single `Parser.parse()` interface (#44267), with the Responses parser migrated to it (#42977).
Removed the stale cuDNN frontend upper bound (#42599); Docker fixes for flashinfer-jit-cache (#44366), FlashInfer CuTe DSL JIT `libcublas-dev` (#39855), and CUTLASS DSL cu13 install order (#45204).
Deprecations
Deprecate `JAISLMHeadModel` (#43784).
Begin the deprecation cycle for the NixlConnector `kv_both` role (#43874).
Remove functions previously scheduled for deprecation in v0.21.0 (#43358).

2026-06-05

No breaking-change keywords detected in release notes.

2026-05-29

⚠ Potential breaking changes

Kernel ABI**: continued migration to libtorch stable ABI — 5/n (#42339), 6/n (#42663), 7/n (#43209).
Experimental**: breakable CUDA graph (#42304).
EPLB: change default EPLB communicator (#43110), VLM-wrapper init fix (#39805), remove dead `torch.accelerator.synchronize()` (#40733).
Deprecations & Removals
Removed old locations of `get_tokenizer` and `resolve_hf_chat_template` (#35024).
Removed deprecated MLA prefill arguments (#42555).
Removed dead CUDA kernels and dead code (#42767, #42889, #43144).

2026-05-15

⚠ Potential breaking changes

Transformers v4 deprecated**: This release formally deprecates `transformers` v4 support (#40389). Users should migrate to `transformers` v5.
C++20 build requirement**: vLLM now requires a C++20-compatible compiler for compatibility with PyTorch (#40380). This is a **breaking build change**.
Breaking**: C++20 required for PyTorch compatibility (#40380).
Breaking**: Transformers v4 deprecated (#40389).

2026-05-10

⚠ Potential breaking changes

Qwen3-VL**: Removed an invalid deepstack boundary check that could fail under heavy load (#40932).

2026-05-04

No breaking-change keywords detected in release notes.

2026-04-27

⚠ Potential breaking changes

PyTorch 2.11 upgrade** (#34644): vLLM ships on torch 2.11 for CUDA, and XPU is now also on torch 2.11 (#37947) — XPU is no longer pinned to 2.10. This is a breaking change for environment dependency.
Deprecations**: Petit NVFP4 removed (#32694).
PyTorch 2.11** for CUDA (#34644) and XPU (#37947) — XPU no longer pinned to 2.10.
Removed / moved**: `resampy` dependency dropped (#39524), `librosa` direct dependency dropped (#39079), `pyav` and `soundfile` moved to common requirements (#39997).
Breaking Changes
3. **Metrics rework**: `vllm:prompt_tokens_recomputed` removed (#38709); `num_cached_tokens` / `num_external_computed_tokens` replaced with `PrefillStats` (#37460).
7. **Petit NVFP4 quantization removed** (#32694); `LLM.reward` deprecated, use `LLM.encode` (#40688); `cprofile` / `cprofile_context` deprecated (#39100); V0 `accept output buffer` deprecated (#39125).
V0 Deprecation