diff --git a/.readthedocs.yaml b/.readthedocs.yaml
index 34735700a224e..284196bc2d279 100644
--- a/.readthedocs.yaml
+++ b/.readthedocs.yaml
@@ -6,7 +6,7 @@ version: 2
 build:
   os: ubuntu-22.04
   tools:
-    python: '3.9'
+    python: "3.12"
 
 sphinx:
   configuration: docs/source/conf.py
diff --git a/CMakeLists.txt b/CMakeLists.txt
index c372ba98befbf..25c0865a90a67 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -31,7 +31,7 @@ install(CODE "set(CMAKE_INSTALL_LOCAL_ONLY TRUE)" ALL_COMPONENTS)
 # Supported python versions.  These versions will be searched in order, the
 # first match will be selected.  These should be kept in sync with setup.py.
 #
-set(PYTHON_SUPPORTED_VERSIONS "3.8" "3.9" "3.10" "3.11" "3.12")
+set(PYTHON_SUPPORTED_VERSIONS "3.9" "3.10" "3.11" "3.12")
 
 # Supported NVIDIA architectures.
 set(CUDA_SUPPORTED_ARCHS "7.0;7.5;8.0;8.6;8.9;9.0")
diff --git a/docs/source/getting_started/amd-installation.rst b/docs/source/getting_started/amd-installation.rst
index 301337aebcf4c..ece5d785e0c65 100644
--- a/docs/source/getting_started/amd-installation.rst
+++ b/docs/source/getting_started/amd-installation.rst
@@ -13,8 +13,6 @@ Requirements
 * GPU: MI200s (gfx90a), MI300 (gfx942), Radeon RX 7900 series (gfx1100)
 * ROCm 6.2
 
-Note: PyTorch 2.5+/ROCm6.2 dropped the support for python 3.8.
-
 Installation options:
 
 #. :ref:`Build from source with docker <build_from_source_docker_rocm>`
diff --git a/docs/source/getting_started/installation.rst b/docs/source/getting_started/installation.rst
index 61871cdf41125..efc050dd1bfb2 100644
--- a/docs/source/getting_started/installation.rst
+++ b/docs/source/getting_started/installation.rst
@@ -66,7 +66,7 @@ If you want to access the wheels for previous commits, you can specify the commi
     $ export VLLM_COMMIT=33f460b17a54acb3b6cc0b03f4a17876cff5eafd # use full commit hash from the main branch
     $ pip install https://vllm-wheels.s3.us-west-2.amazonaws.com/${VLLM_COMMIT}/vllm-1.0.0.dev-cp38-abi3-manylinux1_x86_64.whl
 
-Note that the wheels are built with Python 3.8 ABI (see `PEP 425 <https://peps.python.org/pep-0425/>`_ for more details about ABI), so **they are compatible with Python 3.8 and later**. The version string in the wheel file name (``1.0.0.dev``) is just a placeholder to have a unified URL for the wheels. The actual versions of wheels are contained in the wheel metadata.
+Note that the wheels are built with Python 3.9 ABI (see `PEP 425 <https://peps.python.org/pep-0425/>`_ for more details about ABI), so **they are compatible with Python 3.9 and later**. The version string in the wheel file name (``1.0.0.dev``) is just a placeholder to have a unified URL for the wheels. The actual versions of wheels are contained in the wheel metadata.
 
 Another way to access the latest code is to use the docker images:
 
diff --git a/docs/source/getting_started/neuron-installation.rst b/docs/source/getting_started/neuron-installation.rst
index ec99fc013057b..025ba6ef7ebd8 100644
--- a/docs/source/getting_started/neuron-installation.rst
+++ b/docs/source/getting_started/neuron-installation.rst
@@ -11,7 +11,7 @@ Requirements
 ------------
 
 * OS: Linux
-* Python: 3.8 -- 3.11
+* Python: 3.9 -- 3.11
 * Accelerator: NeuronCore_v2 (in trn1/inf2 instances)
 * Pytorch 2.0.1/2.1.1
 * AWS Neuron SDK 2.16/2.17 (Verified on python 3.8)
diff --git a/docs/source/getting_started/quickstart.rst b/docs/source/getting_started/quickstart.rst
index 00b762ccc2ccb..0c0491c860563 100644
--- a/docs/source/getting_started/quickstart.rst
+++ b/docs/source/getting_started/quickstart.rst
@@ -12,7 +12,7 @@ This guide will help you quickly get started with vLLM to:
 Prerequisites
 --------------
 - OS: Linux
-- Python: 3.8 - 3.12
+- Python: 3.9 -- 3.12
 - GPU: compute capability 7.0 or higher (e.g., V100, T4, RTX20xx, A100, L4, H100, etc.)
 
 Installation
diff --git a/setup.py b/setup.py
index 4a20e49235ac8..d2438ae74c455 100644
--- a/setup.py
+++ b/setup.py
@@ -55,12 +55,6 @@ def is_ninja_available() -> bool:
     return which("ninja") is not None
 
 
-def remove_prefix(text, prefix):
-    if text.startswith(prefix):
-        return text[len(prefix):]
-    return text
-
-
 class CMakeExtension(Extension):
 
     def __init__(self, name: str, cmake_lists_dir: str = '.', **kwa) -> None:
@@ -197,8 +191,10 @@ def build_extensions(self) -> None:
             os.makedirs(self.build_temp)
 
         targets = []
-        target_name = lambda s: remove_prefix(remove_prefix(s, "vllm."),
-                                              "vllm_flash_attn.")
+
+        def target_name(s: str) -> str:
+            return s.removeprefix("vllm.").removeprefix("vllm_flash_attn.")
+
         # Build all the extensions
         for ext in self.extensions:
             self.configure(ext)
diff --git a/vllm/distributed/parallel_state.py b/vllm/distributed/parallel_state.py
index efa3525910a5e..0d15403264eee 100644
--- a/vllm/distributed/parallel_state.py
+++ b/vllm/distributed/parallel_state.py
@@ -89,12 +89,11 @@ def _get_unique_name(name: str) -> str:
     return newname
 
 
-_groups: Dict[str, Callable[[], "GroupCoordinator"]] = {}
+_groups: Dict[str, Callable[[], Optional["GroupCoordinator"]]] = {}
 
 
 def _register_group(group: "GroupCoordinator") -> None:
-    # looks like Python 3.8 does not understand `ReferenceType`
-    _groups[group.unique_name] = weakref.ref(group)  # type: ignore
+    _groups[group.unique_name] = weakref.ref(group)
 
 
 if supports_custom_op():