diff --git a/.github/workflows/python-tests.yml b/.github/workflows/python-tests.yml
index 8006e898e2..fb040d9492 100644
--- a/.github/workflows/python-tests.yml
+++ b/.github/workflows/python-tests.yml
@@ -21,7 +21,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.7", "3.11"]
+        python-version: ["3.8", "3.11"]
         test_name:
           [
             "Repository only",
@@ -29,13 +29,13 @@ jobs:
             "torch",
           ]
         include:
-          - python-version: "3.11" # LFS not ran on 3.7
+          - python-version: "3.11" # LFS not ran on 3.8
             test_name: "lfs"
-          - python-version: "3.7"
+          - python-version: "3.8"
             test_name: "fastai"  # fastai not supported on 3.11 -> test it on 3.10
           - python-version: "3.10"
             test_name: "fastai"
-          - python-version: "3.7"
+          - python-version: "3.8"
             test_name: "tensorflow" # Tensorflow not supported on 3.11 -> test it on 3.10
           - python-version: "3.10"
             test_name: "tensorflow"
@@ -131,14 +131,14 @@ jobs:
   build-windows:
     # (almost) Duplicate config compared to `build-ubuntu` but running on Windows.
     # Please make sure to keep it updated as well.
-    # Lighter version of the tests with only 1 test suite running on Python3.7.
+    # Lighter version of the tests with only 1 test suite running on Python3.8.
     runs-on: windows-latest
     env:
       DISABLE_SYMLINKS_IN_WINDOWS_TESTS: 1
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.7"]
+        python-version: ["3.8"]
 
     steps:
       - uses: actions/checkout@v2
diff --git a/docs/source/installation.md b/docs/source/installation.md
index 423051e10c..1426f20818 100644
--- a/docs/source/installation.md
+++ b/docs/source/installation.md
@@ -6,7 +6,7 @@ rendered properly in your Markdown viewer.
 
 Before you start, you will need to setup your environment by installing the appropriate packages.
 
-`huggingface_hub` is tested on **Python 3.7+**.
+`huggingface_hub` is tested on **Python 3.8+**.
 
 ## Install with pip
 
diff --git a/setup.py b/setup.py
index 1af9002fd0..a16ed1d8dd 100644
--- a/setup.py
+++ b/setup.py
@@ -18,7 +18,6 @@ def get_version() -> str:
     "tqdm>=4.42.1",
     "pyyaml>=5.1",
     "typing-extensions>=3.7.4.3",  # to be able to import TypeAlias
-    "importlib_metadata;python_version<'3.8'",
     "packaging>=20.9",
 ]
 
@@ -108,7 +107,7 @@ def get_version() -> str:
         "console_scripts": ["huggingface-cli=huggingface_hub.commands.huggingface_cli:main"],
         "fsspec.specs": "hf=huggingface_hub.HfFileSystem",
     },
-    python_requires=">=3.7.0",
+    python_requires=">=3.8.0",
     install_requires=install_requires,
     classifiers=[
         "Intended Audience :: Developers",
@@ -118,7 +117,6 @@ def get_version() -> str:
         "Operating System :: OS Independent",
         "Programming Language :: Python :: 3",
         "Programming Language :: Python :: 3 :: Only",
-        "Programming Language :: Python :: 3.7",
         "Programming Language :: Python :: 3.8",
         "Programming Language :: Python :: 3.9",
         "Programming Language :: Python :: 3.10",
diff --git a/src/huggingface_hub/__init__.py b/src/huggingface_hub/__init__.py
index f2d6490d6e..98b3db2bbd 100644
--- a/src/huggingface_hub/__init__.py
+++ b/src/huggingface_hub/__init__.py
@@ -46,7 +46,7 @@
 from typing import TYPE_CHECKING
 
 
-__version__ = "0.16.0.dev0"
+__version__ = "0.17.0.dev0"
 
 # Alphabetical order of definitions is ensured in tests
 # WARNING: any comment added in this dictionary definition will be lost when
diff --git a/src/huggingface_hub/_commit_api.py b/src/huggingface_hub/_commit_api.py
index 7cbcc0d765..ae69085b5a 100644
--- a/src/huggingface_hub/_commit_api.py
+++ b/src/huggingface_hub/_commit_api.py
@@ -10,7 +10,7 @@
 from dataclasses import dataclass, field
 from itertools import groupby
 from pathlib import Path, PurePosixPath
-from typing import TYPE_CHECKING, Any, BinaryIO, Dict, Iterable, Iterator, List, Optional, Tuple, Union
+from typing import TYPE_CHECKING, Any, BinaryIO, Dict, Iterable, Iterator, List, Literal, Optional, Tuple, Union
 
 from tqdm.contrib.concurrent import thread_map
 
@@ -28,7 +28,6 @@
     validate_hf_hub_args,
 )
 from .utils import tqdm as hf_tqdm
-from .utils._typing import Literal
 
 
 if TYPE_CHECKING:
diff --git a/src/huggingface_hub/_snapshot_download.py b/src/huggingface_hub/_snapshot_download.py
index aa85d0f4fe..503ab1c87b 100644
--- a/src/huggingface_hub/_snapshot_download.py
+++ b/src/huggingface_hub/_snapshot_download.py
@@ -1,6 +1,6 @@
 import os
 from pathlib import Path
-from typing import Dict, List, Optional, Union
+from typing import Dict, List, Literal, Optional, Union
 
 from tqdm.auto import tqdm as base_tqdm
 from tqdm.contrib.concurrent import thread_map
@@ -15,7 +15,6 @@
 from .hf_api import HfApi
 from .utils import filter_repo_objects, logging, validate_hf_hub_args
 from .utils import tqdm as hf_tqdm
-from .utils._typing import Literal
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/huggingface_hub/_webhooks_payload.py b/src/huggingface_hub/_webhooks_payload.py
index bc9aa7be04..5d66b99643 100644
--- a/src/huggingface_hub/_webhooks_payload.py
+++ b/src/huggingface_hub/_webhooks_payload.py
@@ -13,12 +13,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """Contains data structures to parse the webhooks payload."""
-from typing import List, Optional
+from typing import List, Literal, Optional
 
 from pydantic import BaseModel
 
-from .utils._typing import Literal
-
 
 # This is an adaptation of the ReportV3 interface implemented in moon-landing. V0, V1 and V2 have been ignored as they
 # are not in used anymore. To keep in sync when format is updated in
diff --git a/src/huggingface_hub/community.py b/src/huggingface_hub/community.py
index cadad72591..f5422ca911 100644
--- a/src/huggingface_hub/community.py
+++ b/src/huggingface_hub/community.py
@@ -6,11 +6,10 @@
 """
 from dataclasses import dataclass
 from datetime import datetime
-from typing import List, Optional
+from typing import List, Literal, Optional
 
 from .constants import REPO_TYPE_MODEL
 from .utils import parse_datetime
-from .utils._typing import Literal
 
 
 DiscussionStatus = Literal["open", "closed", "merged", "draft"]
diff --git a/src/huggingface_hub/file_download.py b/src/huggingface_hub/file_download.py
index c3c7a797c3..fb5beb6a8b 100644
--- a/src/huggingface_hub/file_download.py
+++ b/src/huggingface_hub/file_download.py
@@ -14,7 +14,7 @@
 from functools import partial
 from hashlib import sha256
 from pathlib import Path
-from typing import Any, BinaryIO, Dict, Generator, Optional, Tuple, Union
+from typing import Any, BinaryIO, Dict, Generator, Literal, Optional, Tuple, Union
 from urllib.parse import quote, urlparse
 
 import requests
@@ -64,7 +64,7 @@
 )
 from .utils._headers import _http_user_agent
 from .utils._runtime import _PY_VERSION  # noqa: F401 # for backward compatibility
-from .utils._typing import HTTP_METHOD_T, Literal
+from .utils._typing import HTTP_METHOD_T
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/huggingface_hub/hf_api.py b/src/huggingface_hub/hf_api.py
index 727cbb441e..dab2ceba26 100644
--- a/src/huggingface_hub/hf_api.py
+++ b/src/huggingface_hub/hf_api.py
@@ -26,7 +26,22 @@
 from functools import wraps
 from itertools import islice
 from pathlib import Path
-from typing import Any, BinaryIO, Callable, Dict, Iterable, Iterator, List, Optional, Tuple, TypeVar, Union, overload
+from typing import (
+    Any,
+    BinaryIO,
+    Callable,
+    Dict,
+    Iterable,
+    Iterator,
+    List,
+    Literal,
+    Optional,
+    Tuple,
+    TypedDict,
+    TypeVar,
+    Union,
+    overload,
+)
 from urllib.parse import quote
 
 import requests
@@ -99,7 +114,7 @@
 from .utils._deprecation import (
     _deprecate_arguments,
 )
-from .utils._typing import CallableT, Literal, TypedDict
+from .utils._typing import CallableT
 from .utils.endpoint_helpers import (
     AttributeDictionary,
     DatasetFilter,
diff --git a/src/huggingface_hub/inference/_client.py b/src/huggingface_hub/inference/_client.py
index 868d1cea5a..8434fdf739 100644
--- a/src/huggingface_hub/inference/_client.py
+++ b/src/huggingface_hub/inference/_client.py
@@ -43,6 +43,7 @@
     Dict,
     Iterable,
     List,
+    Literal,
     Optional,
     Union,
     overload,
@@ -80,7 +81,6 @@
     get_session,
     hf_raise_for_status,
 )
-from huggingface_hub.utils._typing import Literal
 
 
 if TYPE_CHECKING:
diff --git a/src/huggingface_hub/inference/_common.py b/src/huggingface_hub/inference/_common.py
index 73c3e61dd8..45385fa41e 100644
--- a/src/huggingface_hub/inference/_common.py
+++ b/src/huggingface_hub/inference/_common.py
@@ -29,6 +29,7 @@
     Generator,
     Iterable,
     List,
+    Literal,
     Optional,
     Set,
     Union,
@@ -46,7 +47,6 @@
     is_numpy_available,
     is_pillow_available,
 )
-from ..utils._typing import Literal
 from ._text_generation import (
     TextGenerationStreamResponse,
 )
diff --git a/src/huggingface_hub/inference/_generated/_async_client.py b/src/huggingface_hub/inference/_generated/_async_client.py
index 7a40403abe..2ca84f1c24 100644
--- a/src/huggingface_hub/inference/_generated/_async_client.py
+++ b/src/huggingface_hub/inference/_generated/_async_client.py
@@ -28,6 +28,7 @@
     AsyncIterable,
     Dict,
     List,
+    Literal,
     Optional,
     Union,
     overload,
@@ -61,7 +62,6 @@
 from huggingface_hub.utils import (
     build_hf_headers,
 )
-from huggingface_hub.utils._typing import Literal
 
 from .._common import _async_yield_from, _import_aiohttp
 
diff --git a/src/huggingface_hub/inference/_types.py b/src/huggingface_hub/inference/_types.py
index b4691a8cc0..5e52dcbc56 100644
--- a/src/huggingface_hub/inference/_types.py
+++ b/src/huggingface_hub/inference/_types.py
@@ -12,9 +12,7 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from typing import TYPE_CHECKING, List
-
-from ..utils._typing import TypedDict
+from typing import TYPE_CHECKING, List, TypedDict
 
 
 if TYPE_CHECKING:
diff --git a/src/huggingface_hub/inference_api.py b/src/huggingface_hub/inference_api.py
index f5d8d5ec8e..1dd3b36924 100644
--- a/src/huggingface_hub/inference_api.py
+++ b/src/huggingface_hub/inference_api.py
@@ -4,6 +4,7 @@
 from .constants import INFERENCE_ENDPOINT
 from .hf_api import HfApi
 from .utils import build_hf_headers, get_session, is_pillow_available, logging, validate_hf_hub_args
+from .utils._deprecation import _deprecate_method
 
 
 logger = logging.get_logger(__name__)
@@ -90,15 +91,14 @@ class InferenceApi:
     """
 
     @validate_hf_hub_args
-    # TODO: add deprecation starting from version v0.16.0 so that we can proactively adapt external scripts.
-    # @_deprecate_method(
-    #     version="0.18.0",
-    #     message=(
-    #         "`InferenceApi` client is deprecated in favor of the more feature-complete `InferenceClient`. Check out"
-    #         " this guide to learn how to convert your script to use it:"
-    #         " https://huggingface.co/docs/huggingface_hub/guides/inference#legacy-inferenceapi-client."
-    #     ),
-    # )
+    @_deprecate_method(
+        version="0.19.0",
+        message=(
+            "`InferenceApi` client is deprecated in favor of the more feature-complete `InferenceClient`. Check out"
+            " this guide to learn how to convert your script to use it:"
+            " https://huggingface.co/docs/huggingface_hub/guides/inference#legacy-inferenceapi-client."
+        ),
+    )
     def __init__(
         self,
         repo_id: str,
diff --git a/src/huggingface_hub/lfs.py b/src/huggingface_hub/lfs.py
index f2474e2a15..c3c890044d 100644
--- a/src/huggingface_hub/lfs.py
+++ b/src/huggingface_hub/lfs.py
@@ -22,7 +22,7 @@
 from math import ceil
 from os.path import getsize
 from pathlib import Path
-from typing import TYPE_CHECKING, BinaryIO, Dict, Iterable, List, Optional, Tuple
+from typing import TYPE_CHECKING, BinaryIO, Dict, Iterable, List, Optional, Tuple, TypedDict
 
 from requests.auth import HTTPBasicAuth
 
@@ -30,7 +30,6 @@
 from huggingface_hub.utils import get_session
 
 from .utils import get_token_to_send, hf_raise_for_status, http_backoff, logging, validate_hf_hub_args
-from .utils._typing import TypedDict
 from .utils.sha import sha256, sha_fileobj
 
 
diff --git a/src/huggingface_hub/repocard.py b/src/huggingface_hub/repocard.py
index 1c4b976ffa..b8e9dc200c 100644
--- a/src/huggingface_hub/repocard.py
+++ b/src/huggingface_hub/repocard.py
@@ -1,7 +1,7 @@
 import os
 import re
 from pathlib import Path
-from typing import Any, Dict, Optional, Type, Union
+from typing import Any, Dict, Literal, Optional, Type, Union
 
 import requests
 import yaml
@@ -21,7 +21,6 @@
 
 from .constants import REPOCARD_NAME
 from .utils import EntryNotFoundError, SoftTemporaryDirectory, validate_hf_hub_args
-from .utils._typing import Literal
 from .utils.logging import get_logger
 
 
diff --git a/src/huggingface_hub/repository.py b/src/huggingface_hub/repository.py
index 757995870b..d13b287aac 100644
--- a/src/huggingface_hub/repository.py
+++ b/src/huggingface_hub/repository.py
@@ -6,7 +6,7 @@
 import time
 from contextlib import contextmanager
 from pathlib import Path
-from typing import Callable, Dict, Iterator, List, Optional, Tuple, Union
+from typing import Callable, Dict, Iterator, List, Optional, Tuple, TypedDict, Union
 from urllib.parse import urlparse
 
 from huggingface_hub.constants import REPO_TYPES_URL_PREFIXES, REPOCARD_NAME
@@ -22,7 +22,6 @@
     tqdm,
     validate_hf_hub_args,
 )
-from .utils._typing import TypedDict
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/huggingface_hub/utils/_cache_manager.py b/src/huggingface_hub/utils/_cache_manager.py
index 3e1443a789..674fe27f30 100644
--- a/src/huggingface_hub/utils/_cache_manager.py
+++ b/src/huggingface_hub/utils/_cache_manager.py
@@ -19,11 +19,10 @@
 from collections import defaultdict
 from dataclasses import dataclass
 from pathlib import Path
-from typing import Dict, FrozenSet, List, Optional, Set, Union
+from typing import Dict, FrozenSet, List, Literal, Optional, Set, Union
 
 from ..constants import HUGGINGFACE_HUB_CACHE
 from . import logging
-from ._typing import Literal
 
 
 logger = logging.get_logger(__name__)
diff --git a/src/huggingface_hub/utils/_http.py b/src/huggingface_hub/utils/_http.py
index 835ef648e9..8d349833e8 100644
--- a/src/huggingface_hub/utils/_http.py
+++ b/src/huggingface_hub/utils/_http.py
@@ -127,7 +127,7 @@ def backend_factory() -> requests.Session:
     return _get_session_from_cache(thread_ident=threading.get_ident())
 
 
-@lru_cache(maxsize=128)  # default value for Python>=3.8. Let's keep the same for Python3.7
+@lru_cache
 def _get_session_from_cache(thread_ident: int) -> requests.Session:
     """
     Create a new session per thread using global factory. Using LRU cache (maxsize 128) to avoid memory leaks when
diff --git a/src/huggingface_hub/utils/_runtime.py b/src/huggingface_hub/utils/_runtime.py
index f5f95cb57c..969a38e6b1 100644
--- a/src/huggingface_hub/utils/_runtime.py
+++ b/src/huggingface_hub/utils/_runtime.py
@@ -13,23 +13,16 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """Check presence of installed packages at runtime."""
+import importlib.metadata
 import platform
 import sys
 from typing import Any, Dict
 
-import packaging.version
-
 from .. import __version__, constants
 
 
 _PY_VERSION: str = sys.version.split()[0].rstrip("+")
 
-if packaging.version.Version(_PY_VERSION) < packaging.version.Version("3.8.0"):
-    import importlib_metadata  # type: ignore
-else:
-    import importlib.metadata as importlib_metadata  # type: ignore
-
-
 _package_versions = {}
 
 _CANDIDATES = {
@@ -65,9 +58,9 @@
     _package_versions[candidate_name] = "N/A"
     for name in package_names:
         try:
-            _package_versions[candidate_name] = importlib_metadata.version(name)
+            _package_versions[candidate_name] = importlib.metadata.version(name)
             break
-        except importlib_metadata.PackageNotFoundError:
+        except importlib.metadata.PackageNotFoundError:
             pass
 
 
diff --git a/src/huggingface_hub/utils/_typing.py b/src/huggingface_hub/utils/_typing.py
index c8885eb1eb..56f2ceed6b 100644
--- a/src/huggingface_hub/utils/_typing.py
+++ b/src/huggingface_hub/utils/_typing.py
@@ -13,15 +13,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """Handle typing imports based on system compatibility."""
-import sys
-from typing import Callable, TypeVar
+from typing import Callable, Literal, TypeVar
 
 
-if sys.version_info >= (3, 8):
-    from typing import Literal, TypedDict
-else:
-    from typing_extensions import Literal, TypedDict  # noqa: F401
-
 HTTP_METHOD_T = Literal["GET", "OPTIONS", "HEAD", "POST", "PUT", "PATCH", "DELETE"]
 
 # type hint meaning "function signature not changed by decorator"
diff --git a/tests/test_inference_api.py b/tests/test_inference_api.py
index 42a4a770f1..ec35013d43 100644
--- a/tests/test_inference_api.py
+++ b/tests/test_inference_api.py
@@ -20,7 +20,7 @@
 from huggingface_hub import hf_hub_download
 from huggingface_hub.inference_api import InferenceApi
 
-from .testing_utils import with_production_testing
+from .testing_utils import expect_deprecation, with_production_testing
 
 
 @with_production_testing
@@ -34,6 +34,7 @@ def setUpClass(cls) -> None:
         cls.image_file = hf_hub_download(repo_id="Narsil/image_dummy", repo_type="dataset", filename="lena.png")
         return super().setUpClass()
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_simple_inference(self):
         api = InferenceApi("bert-base-uncased")
         inputs = "Hi, I think [MASK] is cool"
@@ -45,6 +46,7 @@ def test_simple_inference(self):
         self.assertTrue("sequence" in result)
         self.assertTrue("score" in result)
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_inference_with_params(self):
         api = InferenceApi("typeform/distilbert-base-uncased-mnli")
         inputs = "I bought a device but it is not working and I would like to get reimbursed!"
@@ -54,6 +56,7 @@ def test_inference_with_params(self):
         self.assertTrue("sequence" in result)
         self.assertTrue("scores" in result)
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_inference_with_dict_inputs(self):
         api = InferenceApi("distilbert-base-cased-distilled-squad")
         inputs = {
@@ -65,6 +68,7 @@ def test_inference_with_dict_inputs(self):
         self.assertTrue("score" in result)
         self.assertTrue("answer" in result)
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_inference_with_audio(self):
         api = InferenceApi("facebook/wav2vec2-base-960h")
         file = hf_hub_download(
@@ -77,6 +81,7 @@ def test_inference_with_audio(self):
         self.assertIsInstance(result, dict)
         self.assertTrue("text" in result, f"We received {result} instead")
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_inference_with_image(self):
         api = InferenceApi("google/vit-base-patch16-224")
         data = self.read(self.image_file)
@@ -87,6 +92,7 @@ def test_inference_with_image(self):
             self.assertTrue("score" in classification)
             self.assertTrue("label" in classification)
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_text_to_image(self):
         api = InferenceApi("stabilityai/stable-diffusion-2-1")
         with patch("huggingface_hub.inference_api.get_session") as mock:
@@ -95,6 +101,7 @@ def test_text_to_image(self):
             output = api("cat")
         self.assertIsInstance(output, Image.Image)
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_text_to_image_raw_response(self):
         api = InferenceApi("stabilityai/stable-diffusion-2-1")
         with patch("huggingface_hub.inference_api.get_session") as mock:
@@ -104,6 +111,7 @@ def test_text_to_image_raw_response(self):
         # Raw response is returned
         self.assertEqual(output, mock().post.return_value)
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_inference_overriding_task(self):
         api = InferenceApi(
             "sentence-transformers/paraphrase-albert-small-v2",
@@ -113,10 +121,12 @@ def test_inference_overriding_task(self):
         result = api(inputs)
         self.assertIsInstance(result, list)
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_inference_overriding_invalid_task(self):
         with self.assertRaises(ValueError, msg="Invalid task invalid-task. Make sure it's valid."):
             InferenceApi("bert-base-uncased", task="invalid-task")
 
+    @expect_deprecation("huggingface_hub.inference_api")
     def test_inference_missing_input(self):
         api = InferenceApi("deepset/roberta-base-squad2")
         result = api({"question": "What's my name?"})