diff --git a/src/huggingface_hub/_inference_endpoints.py b/src/huggingface_hub/_inference_endpoints.py
index 651efc97b0..282b627b54 100644
--- a/src/huggingface_hub/_inference_endpoints.py
+++ b/src/huggingface_hub/_inference_endpoints.py
@@ -245,6 +245,7 @@ def update(
         revision: Optional[str] = None,
         task: Optional[str] = None,
         custom_image: Optional[Dict] = None,
+        secrets: Optional[Dict[str, str]] = None,
     ) -> "InferenceEndpoint":
         """Update the Inference Endpoint.
 
@@ -279,7 +280,8 @@ def update(
             custom_image (`Dict`, *optional*):
                 A custom Docker image to use for the Inference Endpoint. This is useful if you want to deploy an
                 Inference Endpoint running on the `text-generation-inference` (TGI) framework (see examples).
-
+            secrets (`Dict[str, str]`, *optional*):
+                Secret values to inject in the container environment.
         Returns:
             [`InferenceEndpoint`]: the same Inference Endpoint, mutated in place with the latest data.
         """
@@ -298,6 +300,7 @@ def update(
             revision=revision,
             task=task,
             custom_image=custom_image,
+            secrets=secrets,
             token=self._token,  # type: ignore [arg-type]
         )
 
diff --git a/src/huggingface_hub/hf_api.py b/src/huggingface_hub/hf_api.py
index 17fa4889a7..3b1e601be5 100644
--- a/src/huggingface_hub/hf_api.py
+++ b/src/huggingface_hub/hf_api.py
@@ -136,9 +136,7 @@
 )
 from .utils import tqdm as hf_tqdm
 from .utils._typing import CallableT
-from .utils.endpoint_helpers import (
-    _is_emission_within_threshold,
-)
+from .utils.endpoint_helpers import _is_emission_within_threshold
 
 
 R = TypeVar("R")  # Return type
@@ -7418,6 +7416,7 @@ def create_inference_endpoint(
         revision: Optional[str] = None,
         task: Optional[str] = None,
         custom_image: Optional[Dict] = None,
+        secrets: Optional[Dict[str, str]] = None,
         type: InferenceEndpointType = InferenceEndpointType.PROTECTED,
         namespace: Optional[str] = None,
         token: Union[bool, str, None] = None,
@@ -7456,6 +7455,8 @@ def create_inference_endpoint(
             custom_image (`Dict`, *optional*):
                 A custom Docker image to use for the Inference Endpoint. This is useful if you want to deploy an
                 Inference Endpoint running on the `text-generation-inference` (TGI) framework (see examples).
+            secrets (`Dict[str, str]`, *optional*):
+                Secret values to inject in the container environment.
             type ([`InferenceEndpointType]`, *optional*):
                 The type of the Inference Endpoint, which can be `"protected"` (default), `"public"` or `"private"`.
             namespace (`str`, *optional*):
@@ -7518,6 +7519,7 @@ def create_inference_endpoint(
             ...         },
             ...         "url": "ghcr.io/huggingface/text-generation-inference:1.1.0",
             ...     },
+            ...    secrets={"MY_SECRET_KEY": "secret_value"},
             ... )
 
             ```
@@ -7543,6 +7545,7 @@ def create_inference_endpoint(
                 "revision": revision,
                 "task": task,
                 "image": image,
+                "secrets": secrets,
             },
             "name": name,
             "provider": {
@@ -7625,6 +7628,7 @@ def update_inference_endpoint(
         revision: Optional[str] = None,
         task: Optional[str] = None,
         custom_image: Optional[Dict] = None,
+        secrets: Optional[Dict[str, str]] = None,
         # Other
         namespace: Optional[str] = None,
         token: Union[bool, str, None] = None,
@@ -7664,7 +7668,8 @@ def update_inference_endpoint(
             custom_image (`Dict`, *optional*):
                 A custom Docker image to use for the Inference Endpoint. This is useful if you want to deploy an
                 Inference Endpoint running on the `text-generation-inference` (TGI) framework (see examples).
-
+            secrets (`Dict[str, str]`, *optional*):
+                Secret values to inject in the container environment.
             namespace (`str`, *optional*):
                 The namespace where the Inference Endpoint will be updated. Defaults to the current user's namespace.
             token (Union[bool, str, None], optional):
@@ -7702,6 +7707,8 @@ def update_inference_endpoint(
             payload["model"]["task"] = task
         if custom_image is not None:
             payload["model"]["image"] = {"custom": custom_image}
+        if secrets is not None:
+            payload["model"]["secrets"] = secrets
 
         response = get_session().put(
             f"{constants.INFERENCE_ENDPOINTS_ENDPOINT}/endpoint/{namespace}/{name}",
diff --git a/tests/test_inference_endpoints.py b/tests/test_inference_endpoints.py
index 019d04f57e..67ec4cab0d 100644
--- a/tests/test_inference_endpoints.py
+++ b/tests/test_inference_endpoints.py
@@ -30,6 +30,7 @@
         "task": "text-generation",
         "framework": "pytorch",
         "image": {"huggingface": {}},
+        "secret": {"token": "my-token"},
     },
     "status": {
         "createdAt": "2023-10-26T12:41:53.263078506Z",
@@ -61,6 +62,7 @@
         "task": "text-generation",
         "framework": "pytorch",
         "image": {"huggingface": {}},
+        "secrets": {"token": "my-token"},
     },
     "status": {
         "createdAt": "2023-10-26T12:41:53.263Z",
@@ -93,6 +95,7 @@
         "task": "text-generation",
         "framework": "pytorch",
         "image": {"huggingface": {}},
+        "secrets": {"token": "my-token"},
     },
     "status": {
         "createdAt": "2023-10-26T12:41:53.263Z",