diff --git a/CHANGES.md b/CHANGES.md
index c16264b8ca4a..a0f2c04b1da0 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -112,6 +112,7 @@
 
 ### Package upgrades
 * [comment]: dependency upgrades (e.g. major version) and migrations that may introduce breaking change
+* (Python) Upgraded the `vllm` extra to vLLM 0.32.0, which requires `openai>=2.25.0`, `transformers>=5.16.1`, and `torch==2.13.0`. The GPU image requirements no longer include vLLM. To install vLLM, use the `vllm` extra ([#41005](https://github.com/apache/beam/issues/41005)).
 
 ### API changes
 * [comment]: Public facing interface changes that may introduce breaking change or change of behaviors
diff --git a/sdks/python/apache_beam/ml/inference/vllm_tests_requirements.txt b/sdks/python/apache_beam/ml/inference/vllm_tests_requirements.txt
index ad5877edbec5..86981cabfcd7 100644
--- a/sdks/python/apache_beam/ml/inference/vllm_tests_requirements.txt
+++ b/sdks/python/apache_beam/ml/inference/vllm_tests_requirements.txt
@@ -17,7 +17,7 @@
 torch>=1.7.1
 torchvision>=0.8.2
 pillow>=8.0.0
-transformers==4.57.1
+transformers>=5.16.1,<5.20.0
 sentencepiece==0.2.1
 google-cloud-monitoring>=2.27.0
-openai>=1.52.2
+openai>=2.25.0,<3.0.0
diff --git a/sdks/python/container/common.gradle b/sdks/python/container/common.gradle
index da2871ab72bb..795997bb950c 100644
--- a/sdks/python/container/common.gradle
+++ b/sdks/python/container/common.gradle
@@ -66,7 +66,8 @@ def generatePythonRequirements = tasks.register("generatePythonRequirements") {
     if ("${project.ext.pythonVersion}" != "3.13" && "${project.ext.pythonVersion}" != "3.14") {
       // GPU requirements not used for any containers directly due to
       // licensing, but can be picked up by customers or other consumers for
-      // use.
+      // use. This list excludes the vllm extra because vLLM pins a torch version
+      // above the cap in the transformers extra. Install vLLM with that extra.
       exec {
         executable 'sh'
         args '-c', "cd ${rootDir} && ${runScriptsPath} " +
@@ -74,7 +75,7 @@ def generatePythonRequirements = tasks.register("generatePythonRequirements") {
                 "${files(configurations.sdkSourceTarball.files).singleFile} " +
                 "gpu_image_requirements.txt " +
                 "container/ml " +
-                "[gcp,dataframe,test,tensorflow,tfrecord,torch,transformers,vllm] " +
+                "[gcp,dataframe,test,tensorflow,tfrecord,torch,transformers] " +
                 "${pipExtraOptions}"
       }
     }
diff --git a/sdks/python/container/ml/py311/gpu_image_requirements.txt b/sdks/python/container/ml/py311/gpu_image_requirements.txt
index 2074c6882e96..66b887e8765b 100644
--- a/sdks/python/container/ml/py311/gpu_image_requirements.txt
+++ b/sdks/python/container/ml/py311/gpu_image_requirements.txt
@@ -22,7 +22,6 @@
 # Reach out to a committer if you need help.
 
 absl-py==2.5.1
-agent-detector==2.0.0
 aiohappyeyeballs==2.7.1
 aiohttp==3.14.5
 aiosignal==1.4.0
@@ -31,49 +30,32 @@ annotated-types==0.8.0
 anyio==4.16.0
 arro3-core==0.9.1
 asn1crypto==1.5.1
-astor==0.8.1
 astunparse==1.6.3
 attrs==26.1.0
 backports.tarfile==1.2.0
 beartype==0.22.9
 beautifulsoup4==4.15.0
-blake3==1.0.11
 bs4==0.0.2
 build==1.6.1
 cachetools==6.2.6
-cbor2==6.2.0.post2
 certifi==2026.10.23
 cffi==2.1.1
 charset-normalizer==3.5.2
 click==8.5.0
 cloud-sql-python-connector==1.23.0
 cloudpickle==3.1.2
-compressed-tensors==0.10.2
 crcmod==1.7
 cryptography==50.0.2
-cuda-bindings==13.4.4
-cuda-pathfinder==1.8.5
-cuda-toolkit==13.2.2
-cupy-cuda12x==14.2.0
 Cython==3.3.1
 defusedxml==0.7.1
 deltalake==1.6.6
 Deprecated==1.3.1
-depyf==0.19.0
-detect-installer==0.2.1
 dill==0.3.1.1
-diskcache==5.6.3
 distro==1.9.0
 dnspython==2.9.1
 docker==7.2.0
 docstring_parser==0.18.0
-einops==0.8.2
-email-validator==2.3.0
 execnet==2.1.2
-fastapi==0.162.0
-fastapi-cli==0.0.32
-fastapi-cloud-cli==0.26.0
-fastar==0.12.0
 fastavro==1.13.1
 fasteners==0.20
 filelock==4.2.7
@@ -83,7 +65,6 @@ frozenlist==1.8.0
 fsspec==2026.9.0
 future==1.0.0
 gast==0.7.0
-gguf==0.19.0
 google-api-core==2.43.2
 google-api-python-client==2.201.0
 google-apitools==0.5.31
@@ -129,7 +110,6 @@ hpack==4.2.0
 httpcore==1.0.9
 httpcore2==2.16.0
 httplib2==0.32.1
-httptools==0.9.0
 httpx==0.28.1
 httpx2==2.16.0
 huggingface_hub==2.4.0
@@ -138,86 +118,50 @@ hypothesis==6.148.3
 idna==3.20
 importlib_metadata==10.1.0
 iniconfig==2.3.1
-interegular==0.3.3
 jaraco.classes==3.4.0
 jaraco.context==6.1.2
 jaraco.functools==4.6.0
 jeepney==0.9.0
 Jinja2==3.1.6
-jiter==0.18.0
 joblib==1.6.0
 jsonpickle==4.1.4
-jsonschema==4.26.0
-jsonschema-specifications==2025.9.1
 keras==3.15.1
 keyring==25.7.0
 keyrings.google-artifactregistry-auth==1.1.2
-lark==1.2.2
 libclang==18.1.1
 linkify-it-py==2.2.0
-llguidance==0.7.30
-llvmlite==0.44.0
-lm-format-enforcer==0.10.12
 markdown-it-py==4.2.0
 MarkupSafe==3.0.4
 mdit-py-plugins==0.6.1
 mdurl==0.1.2
 memray==1.19.3
-mistral_common==1.13.1
 ml_dtypes==0.6.0
 mmh3==5.3.1
 mock==5.2.0
 more-itertools==11.2.1
 mpmath==1.3.0
-msgpack==1.2.4
-msgspec==0.22.0
 multidict==7.2.1
 namex==0.1.0
 networkx==3.6.1
-ninja==1.13.2
 nltk==3.10.4
-numba==0.61.2
 numpy==2.2.6
-nvidia-cublas==13.4.1.3
 nvidia-cublas-cu12==12.6.4.1
-nvidia-cuda-cccl==13.3.4.3.1
-nvidia-cuda-cupti==13.2.86
 nvidia-cuda-cupti-cu12==12.6.80
-nvidia-cuda-nvrtc==13.2.86
 nvidia-cuda-nvrtc-cu12==12.6.77
-nvidia-cuda-runtime==13.2.86
 nvidia-cuda-runtime-cu12==12.6.77
 nvidia-cudnn-cu12==9.5.1.17
-nvidia-cudnn-cu13==9.26.0.51
-nvidia-cufft==12.2.0.57
 nvidia-cufft-cu12==11.3.0.4
-nvidia-cufile==1.17.1.22
 nvidia-cufile-cu12==1.11.1.6
 nvidia-curand-cu12==10.3.7.77
-nvidia-cusolver==12.2.0.11
 nvidia-cusolver-cu12==11.7.1.2
-nvidia-cusparse==12.7.10.12
 nvidia-cusparse-cu12==12.5.4.2
 nvidia-cusparselt-cu12==0.6.3
-nvidia-cusparselt-cu13==0.8.1
 nvidia-nccl-cu12==2.26.2
-nvidia-nccl-cu13==2.30.7
-nvidia-nvjitlink==13.4.92
 nvidia-nvjitlink-cu12==12.6.85
-nvidia-nvshmem-cu13==3.7.2
-nvidia-nvtx==13.2.86
 nvidia-nvtx-cu12==12.6.77
 oauth2client==4.1.3
 objsize==0.7.1
-openai==1.107.1
-openai-harmony==0.0.8
-opencv-python-headless==5.0.0.93
 opentelemetry-api==1.45.2
-opentelemetry-exporter-http-transport==0.66b2
-opentelemetry-exporter-otlp-common==0.66b2
-opentelemetry-exporter-otlp-proto-common==1.45.2
-opentelemetry-exporter-otlp-proto-http==1.45.2
-opentelemetry-proto==1.45.2
 opentelemetry-resourcedetector-gcp==1.14.0
 opentelemetry-sdk==1.45.2
 opentelemetry-semantic-conventions==0.66b2
@@ -225,34 +169,24 @@ opt_einsum==3.4.0
 optree==0.20.0
 oracledb==26.1.0
 orjson==3.13.0
-outlines_core==0.2.10
 packaging==26.3
 pandas==2.3.3
 parameterized==0.9.0
-partial-json-parser==0.2.1.1.post7
 pg8000==1.31.6
 pillow==13.0.0
 pip==26.3
 platformdirs==4.14.3
 pluggy==1.7.0
 portalocker==4.4.2
-prometheus-fastapi-instrumentator==8.1.0
-prometheus_client==0.26.0
 propcache==0.5.4
 proto-plus==1.29.0
 protobuf==6.33.6
-psutil==7.2.2
 psycopg2-binary==2.9.13
-py-cpuinfo==9.0.0
 pyarrow==25.0.1
 pyasn1==0.6.5
 pyasn1_modules==0.4.2
-pybase64==1.5.1
-pycountry==26.2.16
 pycparser==3.12
 pydantic==2.14.2
-pydantic-extra-types==2.11.1
-pydantic-settings==2.16.2
 pydantic_core==2.50.2
 pydot==1.4.2
 Pygments==2.21.0
@@ -267,43 +201,29 @@ pytest-timeout==2.4.0
 pytest-xdist==3.8.0
 python-dateutil==2.9.0.post0
 python-dotenv==1.2.4
-python-json-logger==4.2.0
-python-multipart==0.0.33
 python-tds==1.17.1
 pytz==2026.6
 PyYAML==6.0.3
-pyzmq==27.2.0
 qdrant-client==1.20.0
-ray==2.59.0
-referencing==0.37.0
 regex==2026.9.29
 requests==2.34.2
 requests-mock==1.12.1
 rich==15.0.0
-rich-toolkit==0.20.6
-rignore==0.8.1
-rpds-py==2026.9.1
 rsa==4.9.1
 safetensors==0.8.0
 scikit-learn==1.7.2
 scipy==1.17.1
 scramp==1.5.0
 SecretStorage==3.5.0
-sentencepiece==0.2.2
-sentry-sdk==2.73.0
-setproctitle==1.3.8
 setuptools==84.2.1
 shellingham==1.5.4
 six==1.17.0
 sniffio==1.3.1
 sortedcontainers==2.4.0
-soundfile==0.14.0
 soupsieve==3.0.4
-soxr==1.1.0
 SQLAlchemy==2.1.6
 sqlalchemy_pytds==1.0.3
 sqlparse==0.6.1
-starlette==1.10.0
 sympy==1.14.0
 tenacity==9.1.4
 tensorflow==2.21.0
@@ -312,11 +232,8 @@ termcolor==3.3.0
 testcontainers==4.15.0
 textual==8.2.8
 threadpoolctl==3.7.0
-tiktoken==0.14.0
 tokenizers==0.23.3
 torch==2.7.1
-torchaudio==2.7.1
-torchvision==0.22.1
 tqdm==4.70.1
 transformers==5.20.2
 triton==3.3.1
@@ -327,16 +244,10 @@ typing_extensions==4.16.0
 tzdata==2026.6
 uritemplate==4.2.0
 urllib3==2.8.0
-uvicorn==0.54.0
-uvloop==0.23.0
 virtualenv-clone==0.5.7
-vllm==0.10.1.1
-watchfiles==1.3.0
 websockets==17.4
 wheel==0.48.0
 wrapt==2.5.2
-xformers==0.0.31
-xgrammar==0.1.21
 yarl==1.25.1
 zipp==4.1.2
 zstandard==0.25.0
diff --git a/sdks/python/container/ml/py312/gpu_image_requirements.txt b/sdks/python/container/ml/py312/gpu_image_requirements.txt
index fee38bf598d2..12ef63ea17c9 100644
--- a/sdks/python/container/ml/py312/gpu_image_requirements.txt
+++ b/sdks/python/container/ml/py312/gpu_image_requirements.txt
@@ -22,7 +22,6 @@
 # Reach out to a committer if you need help.
 
 absl-py==2.5.1
-agent-detector==2.0.0
 aiohappyeyeballs==2.7.1
 aiohttp==3.14.5
 aiosignal==1.4.0
@@ -31,48 +30,31 @@ annotated-types==0.8.0
 anyio==4.16.0
 arro3-core==0.9.1
 asn1crypto==1.5.1
-astor==0.8.1
 astunparse==1.6.3
 attrs==26.1.0
 beartype==0.22.9
 beautifulsoup4==4.15.0
-blake3==1.0.11
 bs4==0.0.2
 build==1.6.1
 cachetools==6.2.6
-cbor2==6.2.0.post2
 certifi==2026.10.23
 cffi==2.1.1
 charset-normalizer==3.5.2
 click==8.5.0
 cloud-sql-python-connector==1.23.0
 cloudpickle==3.1.2
-compressed-tensors==0.10.2
 crcmod==1.7
 cryptography==50.0.2
-cuda-bindings==13.4.4
-cuda-pathfinder==1.8.5
-cuda-toolkit==13.2.2
-cupy-cuda12x==14.2.0
 Cython==3.3.1
 defusedxml==0.7.1
 deltalake==1.6.6
 Deprecated==3.0.1
-depyf==0.19.0
-detect-installer==0.2.1
 dill==0.3.1.1
-diskcache==5.6.3
 distro==1.9.0
 dnspython==2.9.1
 docker==7.2.0
 docstring_parser==0.18.0
-einops==0.8.2
-email-validator==2.3.0
 execnet==2.1.2
-fastapi==0.162.0
-fastapi-cli==0.0.32
-fastapi-cloud-cli==0.26.0
-fastar==0.12.0
 fastavro==1.13.1
 fasteners==0.20
 filelock==4.2.7
@@ -82,7 +64,6 @@ frozenlist==1.8.0
 fsspec==2026.9.0
 future==1.0.0
 gast==0.7.0
-gguf==0.19.0
 google-api-core==2.43.2
 google-api-python-client==2.201.0
 google-apitools==0.5.31
@@ -128,7 +109,6 @@ hpack==4.2.0
 httpcore==1.0.9
 httpcore2==2.16.0
 httplib2==0.32.1
-httptools==0.9.0
 httpx==0.28.1
 httpx2==2.16.0
 huggingface_hub==2.4.0
@@ -136,86 +116,50 @@ hyperframe==6.1.0
 hypothesis==6.148.3
 idna==3.20
 iniconfig==2.3.1
-interegular==0.3.3
 jaraco.classes==3.4.0
 jaraco.context==6.1.2
 jaraco.functools==4.6.0
 jeepney==0.9.0
 Jinja2==3.1.6
-jiter==0.18.0
 joblib==1.6.0
 jsonpickle==4.1.4
-jsonschema==4.26.0
-jsonschema-specifications==2025.9.1
 keras==3.15.1
 keyring==25.7.0
 keyrings.google-artifactregistry-auth==1.1.2
-lark==1.2.2
 libclang==18.1.1
 linkify-it-py==2.2.0
-llguidance==0.7.30
-llvmlite==0.44.0
-lm-format-enforcer==0.10.12
 markdown-it-py==4.2.0
 MarkupSafe==3.0.4
 mdit-py-plugins==0.6.1
 mdurl==0.1.2
 memray==1.19.3
-mistral_common==1.13.1
 ml_dtypes==0.6.0
 mmh3==5.3.1
 mock==5.2.0
 more-itertools==11.2.1
 mpmath==1.3.0
-msgpack==1.2.4
-msgspec==0.22.0
 multidict==7.2.1
 namex==0.1.0
 networkx==3.7
-ninja==1.13.2
 nltk==3.10.4
-numba==0.61.2
 numpy==2.2.6
-nvidia-cublas==13.4.1.3
 nvidia-cublas-cu12==12.6.4.1
-nvidia-cuda-cccl==13.3.4.3.1
-nvidia-cuda-cupti==13.2.86
 nvidia-cuda-cupti-cu12==12.6.80
-nvidia-cuda-nvrtc==13.2.86
 nvidia-cuda-nvrtc-cu12==12.6.77
-nvidia-cuda-runtime==13.2.86
 nvidia-cuda-runtime-cu12==12.6.77
 nvidia-cudnn-cu12==9.5.1.17
-nvidia-cudnn-cu13==9.26.0.51
-nvidia-cufft==12.2.0.57
 nvidia-cufft-cu12==11.3.0.4
-nvidia-cufile==1.17.1.22
 nvidia-cufile-cu12==1.11.1.6
 nvidia-curand-cu12==10.3.7.77
-nvidia-cusolver==12.2.0.11
 nvidia-cusolver-cu12==11.7.1.2
-nvidia-cusparse==12.7.10.12
 nvidia-cusparse-cu12==12.5.4.2
 nvidia-cusparselt-cu12==0.6.3
-nvidia-cusparselt-cu13==0.8.1
 nvidia-nccl-cu12==2.26.2
-nvidia-nccl-cu13==2.30.7
-nvidia-nvjitlink==13.4.92
 nvidia-nvjitlink-cu12==12.6.85
-nvidia-nvshmem-cu13==3.7.2
-nvidia-nvtx==13.2.86
 nvidia-nvtx-cu12==12.6.77
 oauth2client==4.1.3
 objsize==0.7.1
-openai==1.107.1
-openai-harmony==0.0.8
-opencv-python-headless==5.0.0.93
 opentelemetry-api==1.45.2
-opentelemetry-exporter-http-transport==0.66b2
-opentelemetry-exporter-otlp-common==0.66b2
-opentelemetry-exporter-otlp-proto-common==1.45.2
-opentelemetry-exporter-otlp-proto-http==1.45.2
-opentelemetry-proto==1.45.2
 opentelemetry-resourcedetector-gcp==1.14.0
 opentelemetry-sdk==1.45.2
 opentelemetry-semantic-conventions==0.66b2
@@ -223,34 +167,24 @@ opt_einsum==3.4.0
 optree==0.20.0
 oracledb==26.1.0
 orjson==3.13.0
-outlines_core==0.2.10
 packaging==26.3
 pandas==2.3.3
 parameterized==0.9.0
-partial-json-parser==0.2.1.1.post7
 pg8000==1.31.6
 pillow==13.0.0
 pip==26.3
 platformdirs==4.14.3
 pluggy==1.7.0
 portalocker==4.4.2
-prometheus-fastapi-instrumentator==8.1.0
-prometheus_client==0.26.0
 propcache==0.5.4
 proto-plus==1.29.0
 protobuf==6.33.6
-psutil==7.2.2
 psycopg2-binary==2.9.13
-py-cpuinfo==9.0.0
 pyarrow==25.0.1
 pyasn1==0.6.5
 pyasn1_modules==0.4.2
-pybase64==1.5.1
-pycountry==26.2.16
 pycparser==3.12
 pydantic==2.14.2
-pydantic-extra-types==2.11.1
-pydantic-settings==2.16.2
 pydantic_core==2.50.2
 pydot==1.4.2
 Pygments==2.21.0
@@ -265,43 +199,29 @@ pytest-timeout==2.4.0
 pytest-xdist==3.8.0
 python-dateutil==2.9.0.post0
 python-dotenv==1.2.4
-python-json-logger==4.2.0
-python-multipart==0.0.33
 python-tds==1.17.1
 pytz==2026.6
 PyYAML==6.0.3
-pyzmq==27.2.0
 qdrant-client==1.20.0
-ray==2.59.0
-referencing==0.37.0
 regex==2026.9.29
 requests==2.34.2
 requests-mock==1.12.1
 rich==15.0.0
-rich-toolkit==0.20.6
-rignore==0.8.1
-rpds-py==2026.9.1
 rsa==4.9.1
 safetensors==0.8.0
 scikit-learn==1.7.2
 scipy==1.18.1
 scramp==1.5.0
 SecretStorage==3.5.0
-sentencepiece==0.2.2
-sentry-sdk==2.73.0
-setproctitle==1.3.8
 setuptools==79.0.1
 shellingham==1.5.4
 six==1.17.0
 sniffio==1.3.1
 sortedcontainers==2.4.0
-soundfile==0.14.0
 soupsieve==3.0.4
-soxr==1.1.0
 SQLAlchemy==2.1.6
 sqlalchemy_pytds==1.0.3
 sqlparse==0.6.1
-starlette==1.10.0
 sympy==1.14.0
 tenacity==9.1.4
 tensorflow==2.21.0
@@ -310,11 +230,8 @@ termcolor==3.3.0
 testcontainers==4.15.0
 textual==8.2.8
 threadpoolctl==3.7.0
-tiktoken==0.14.0
 tokenizers==0.23.3
 torch==2.7.1
-torchaudio==2.7.1
-torchvision==0.22.1
 tqdm==4.70.1
 transformers==5.20.2
 triton==3.3.1
@@ -325,15 +242,9 @@ typing_extensions==4.16.0
 tzdata==2026.6
 uritemplate==4.2.0
 urllib3==2.8.0
-uvicorn==0.54.0
-uvloop==0.23.0
 virtualenv-clone==0.5.7
-vllm==0.10.1.1
-watchfiles==1.3.0
 websockets==17.4
 wheel==0.48.0
 wrapt==2.5.2
-xformers==0.0.31
-xgrammar==0.1.21
 yarl==1.25.1
 zstandard==0.25.0
diff --git a/sdks/python/setup.py b/sdks/python/setup.py
index 9f9618e30fbc..b289cd6b80f2 100644
--- a/sdks/python/setup.py
+++ b/sdks/python/setup.py
@@ -722,7 +722,7 @@ def get_portability_package_data():
           ],
           'milvus': milvus_dependency,
           'qdrant': qdrant_dependency,
-          'vllm': ['openai==1.107.1', 'vllm==0.10.1.1', 'triton==3.3.1']
+          'vllm': ['openai>=2.25.0,<3.0.0', 'vllm==0.32.0']
       },
       zip_safe=False,
       # PyPI package information.