diff --git a/inference_gateway/settings.py b/inference_gateway/settings.py
index 9e532bf7..876d49bb 100644
--- a/inference_gateway/settings.py
+++ b/inference_gateway/settings.py
@@ -127,6 +127,13 @@
 # Load maintenance notices to be displayed for individual clusters
 MAINTENANCE_ERROR_NOTICES = json.loads(os.getenv("MAINTENANCE_ERROR_NOTICES", "{}"))
 
+# Endpoint slug mapping to reroute legacy slugs to their latest value
+LEGACY_ENDPOINT_SLUG_MAPPING = json.loads(
+    os.getenv("LEGACY_ENDPOINT_SLUG_MAPPING", "{}")
+)
+for key, value in LEGACY_ENDPOINT_SLUG_MAPPING.items():
+    LEGACY_ENDPOINT_SLUG_MAPPING[key] = str(value).strip()
+
 # Allowed hosts
 ALLOWED_HOSTS = textfield_to_strlist(os.getenv("ALLOWED_HOSTS", "*"))
 
diff --git a/resource_server_async/endpoints/endpoint.py b/resource_server_async/endpoints/endpoint.py
index fe3c8ea4..22487d70 100644
--- a/resource_server_async/endpoints/endpoint.py
+++ b/resource_server_async/endpoints/endpoint.py
@@ -7,13 +7,14 @@
 from django.forms.models import model_to_dict
 from django.utils.text import slugify
 
-from inference_gateway.settings import MODEL_DETAILS_KEYS
+from inference_gateway.settings import LEGACY_ENDPOINT_SLUG_MAPPING, MODEL_DETAILS_KEYS
 from resource_server_async.cache import get_redis_client
 from resource_server_async.rate_limiters import TokenLimiterCheck, TokenRateLimiter
 
 from ..auth import check_permission as auth_utils_check_permission
 from ..errors import (
     BatchUnavailable,
+    EndpointError,
     EndpointNotFound,
     Unauthorized,
 )
@@ -196,6 +197,8 @@ def build_token_limiter(
     async def load_adapter(cls, cluster: str, framework: str, model: str) -> Self:
         """Extract the endpoint from the database and return its underlying adapter object."""
         endpoint_slug = slugify(f"{cluster} {framework} {model.lower()}")
+        if endpoint_slug in LEGACY_ENDPOINT_SLUG_MAPPING:
+            endpoint_slug = LEGACY_ENDPOINT_SLUG_MAPPING[endpoint_slug]
 
         if (adapter := _adapter_cache.get(endpoint_slug)) is not None:
             assert isinstance(adapter, cls)
@@ -229,6 +232,10 @@ async def load_adapter(cls, cluster: str, framework: str, model: str) -> Self:
             raise AssertionError(
                 f"Endpoint adapter {db_endpoint.endpoint_adapter} is not an instance of {cls.__name__}"
             )
+        if endpoint.cluster != cluster or endpoint.framework != framework:
+            raise EndpointError(
+                "Endpoint mapping cannot alter 'cluster' or 'framework'."
+            )
         _adapter_cache[endpoint_slug] = endpoint
         return endpoint
 
diff --git a/resource_server_async/services.py b/resource_server_async/services.py
index bad417b5..2d36fb90 100644
--- a/resource_server_async/services.py
+++ b/resource_server_async/services.py
@@ -288,6 +288,9 @@ async def submit_openai_inference_request(
         f"endpoint_slug: {endpoint.endpoint_slug} - user: {context.user.username}"
     )
 
+    # Overwrite model name in case it has been updated via endpoint slug mapping
+    payload.model = endpoint.model
+
     # Block access if the user is not allowed to use the endpoint
     endpoint.check_permission(context.user)