diff --git a/llama_stack/providers/inline/inference/meta_reference/inference.py b/llama_stack/providers/inline/inference/meta_reference/inference.py
index 1e668b183..844cf6939 100644
--- a/llama_stack/providers/inline/inference/meta_reference/inference.py
+++ b/llama_stack/providers/inline/inference/meta_reference/inference.py
@@ -11,9 +11,11 @@ from typing import AsyncGenerator, List
 from llama_models.sku_list import resolve_model
 
 from llama_models.llama3.api.datatypes import *  # noqa: F403
-from llama_stack.apis.inference import *  # noqa: F403
-from llama_stack.providers.datatypes import Model, ModelsProtocolPrivate
 
+from llama_stack.providers.utils.inference.model_registry import build_model_alias
+from llama_stack.apis.inference import *  # noqa: F403
+from llama_stack.providers.datatypes import ModelsProtocolPrivate
+from llama_stack.providers.utils.inference.model_registry import ModelRegistryHelper
 from llama_stack.providers.utils.inference.prompt_adapter import (
     convert_image_media_to_url,
     request_has_media,
@@ -28,10 +30,19 @@ from .model_parallel import LlamaModelParallelGenerator
 SEMAPHORE = asyncio.Semaphore(1)
 
 
-class MetaReferenceInferenceImpl(Inference, ModelsProtocolPrivate):
+class MetaReferenceInferenceImpl(Inference, ModelRegistryHelper, ModelsProtocolPrivate):
     def __init__(self, config: MetaReferenceInferenceConfig) -> None:
         self.config = config
         model = resolve_model(config.model)
+        ModelRegistryHelper.__init__(
+            self,
+            [
+                build_model_alias(
+                    model.descriptor(),
+                    model.core_model_id,
+                )
+            ],
+        )
         if model is None:
             raise RuntimeError(f"Unknown model: {config.model}, Run `llama model list`")
         self.model = model
diff --git a/llama_stack/providers/remote/inference/bedrock/bedrock.py b/llama_stack/providers/remote/inference/bedrock/bedrock.py
index 47abff689..8762a6c95 100644
--- a/llama_stack/providers/remote/inference/bedrock/bedrock.py
+++ b/llama_stack/providers/remote/inference/bedrock/bedrock.py
@@ -13,7 +13,7 @@ from llama_models.llama3.api.chat_format import ChatFormat
 from llama_models.llama3.api.tokenizer import Tokenizer
 
 from llama_stack.providers.utils.inference.model_registry import (
-    ModelAlias,
+    build_model_alias,
     ModelRegistryHelper,
 )
 
@@ -24,20 +24,17 @@ from llama_stack.providers.utils.bedrock.client import create_bedrock_client
 
 
 model_aliases = [
-    ModelAlias(
-        provider_model_id="meta.llama3-1-8b-instruct-v1:0",
-        aliases=["Llama3.1-8B-Instruct"],
-        llama_model=CoreModelId.llama3_1_8b_instruct,
+    build_model_alias(
+        "meta.llama3-1-8b-instruct-v1:0",
+        CoreModelId.llama3_1_8b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta.llama3-1-70b-instruct-v1:0",
-        aliases=["Llama3.1-70B-Instruct"],
-        llama_model=CoreModelId.llama3_1_70b_instruct,
+    build_model_alias(
+        "meta.llama3-1-70b-instruct-v1:0",
+        CoreModelId.llama3_1_70b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta.llama3-1-405b-instruct-v1:0",
-        aliases=["Llama3.1-405B-Instruct"],
-        llama_model=CoreModelId.llama3_1_405b_instruct,
+    build_model_alias(
+        "meta.llama3-1-405b-instruct-v1:0",
+        CoreModelId.llama3_1_405b_instruct,
     ),
 ]
 
diff --git a/llama_stack/providers/remote/inference/databricks/databricks.py b/llama_stack/providers/remote/inference/databricks/databricks.py
index fedea0f86..1337b6c09 100644
--- a/llama_stack/providers/remote/inference/databricks/databricks.py
+++ b/llama_stack/providers/remote/inference/databricks/databricks.py
@@ -18,7 +18,7 @@ from openai import OpenAI
 from llama_stack.apis.inference import *  # noqa: F403
 
 from llama_stack.providers.utils.inference.model_registry import (
-    ModelAlias,
+    build_model_alias,
     ModelRegistryHelper,
 )
 from llama_stack.providers.utils.inference.openai_compat import (
@@ -34,15 +34,13 @@ from .config import DatabricksImplConfig
 
 
 model_aliases = [
-    ModelAlias(
-        provider_model_id="databricks-meta-llama-3-1-70b-instruct",
-        aliases=["Llama3.1-70B-Instruct"],
-        llama_model=CoreModelId.llama3_1_70b_instruct.value,
+    build_model_alias(
+        "databricks-meta-llama-3-1-70b-instruct",
+        CoreModelId.llama3_1_70b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="databricks-meta-llama-3-1-405b-instruct",
-        aliases=["Llama3.1-405B-Instruct"],
-        llama_model=CoreModelId.llama3_1_405b_instruct.value,
+    build_model_alias(
+        "databricks-meta-llama-3-1-405b-instruct",
+        CoreModelId.llama3_1_405b_instruct,
     ),
 ]
 
diff --git a/llama_stack/providers/remote/inference/fireworks/fireworks.py b/llama_stack/providers/remote/inference/fireworks/fireworks.py
index ce9639cbd..e0d42c721 100644
--- a/llama_stack/providers/remote/inference/fireworks/fireworks.py
+++ b/llama_stack/providers/remote/inference/fireworks/fireworks.py
@@ -15,7 +15,7 @@ from llama_models.llama3.api.tokenizer import Tokenizer
 from llama_stack.apis.inference import *  # noqa: F403
 from llama_stack.distribution.request_headers import NeedsRequestProviderData
 from llama_stack.providers.utils.inference.model_registry import (
-    ModelAlias,
+    build_model_alias,
     ModelRegistryHelper,
 )
 from llama_stack.providers.utils.inference.openai_compat import (
@@ -36,50 +36,41 @@ from .config import FireworksImplConfig
 
 
 model_aliases = [
-    ModelAlias(
-        provider_model_id="fireworks/llama-v3p1-8b-instruct",
-        aliases=["Llama3.1-8B-Instruct"],
-        llama_model=CoreModelId.llama3_1_8b_instruct.value,
+    build_model_alias(
+        "fireworks/llama-v3p1-8b-instruct",
+        CoreModelId.llama3_1_8b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-v3p1-70b-instruct",
-        aliases=["Llama3.1-70B-Instruct"],
-        llama_model=CoreModelId.llama3_1_70b_instruct.value,
+    build_model_alias(
+        "fireworks/llama-v3p1-70b-instruct",
+        CoreModelId.llama3_1_70b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-v3p1-405b-instruct",
-        aliases=["Llama3.1-405B-Instruct"],
-        llama_model=CoreModelId.llama3_1_405b_instruct.value,
+    build_model_alias(
+        "fireworks/llama-v3p1-405b-instruct",
+        CoreModelId.llama3_1_405b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-v3p2-1b-instruct",
-        aliases=["Llama3.2-1B-Instruct"],
-        llama_model=CoreModelId.llama3_2_3b_instruct.value,
+    build_model_alias(
+        "fireworks/llama-v3p2-1b-instruct",
+        CoreModelId.llama3_2_3b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-v3p2-3b-instruct",
-        aliases=["Llama3.2-3B-Instruct"],
-        llama_model=CoreModelId.llama3_2_11b_vision_instruct.value,
+    build_model_alias(
+        "fireworks/llama-v3p2-3b-instruct",
+        CoreModelId.llama3_2_11b_vision_instruct,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-v3p2-11b-vision-instruct",
-        aliases=["Llama3.2-11B-Vision-Instruct"],
-        llama_model=CoreModelId.llama3_2_11b_vision_instruct.value,
+    build_model_alias(
+        "fireworks/llama-v3p2-11b-vision-instruct",
+        CoreModelId.llama3_2_11b_vision_instruct,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-v3p2-90b-vision-instruct",
-        aliases=["Llama3.2-90B-Vision-Instruct"],
-        llama_model=CoreModelId.llama3_2_90b_vision_instruct.value,
+    build_model_alias(
+        "fireworks/llama-v3p2-90b-vision-instruct",
+        CoreModelId.llama3_2_90b_vision_instruct,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-guard-3-8b",
-        aliases=["Llama-Guard-3-8B"],
-        llama_model=CoreModelId.llama_guard_3_8b.value,
+    build_model_alias(
+        "fireworks/llama-guard-3-8b",
+        CoreModelId.llama_guard_3_8b,
     ),
-    ModelAlias(
-        provider_model_id="fireworks/llama-guard-3-11b-vision",
-        aliases=["Llama-Guard-3-11B-Vision"],
-        llama_model=CoreModelId.llama_guard_3_11b_vision.value,
+    build_model_alias(
+        "fireworks/llama-guard-3-11b-vision",
+        CoreModelId.llama_guard_3_11b_vision,
     ),
 ]
 
diff --git a/llama_stack/providers/remote/inference/ollama/ollama.py b/llama_stack/providers/remote/inference/ollama/ollama.py
index 4a7f548a6..34af95b50 100644
--- a/llama_stack/providers/remote/inference/ollama/ollama.py
+++ b/llama_stack/providers/remote/inference/ollama/ollama.py
@@ -15,7 +15,7 @@ from llama_models.llama3.api.tokenizer import Tokenizer
 from ollama import AsyncClient
 
 from llama_stack.providers.utils.inference.model_registry import (
-    ModelAlias,
+    build_model_alias,
     ModelRegistryHelper,
 )
 
@@ -40,40 +40,33 @@ from llama_stack.providers.utils.inference.prompt_adapter import (
 
 
 model_aliases = [
-    ModelAlias(
-        provider_model_id="llama3.1:8b-instruct-fp16",
-        aliases=["Llama3.1-8B-Instruct"],
-        llama_model=CoreModelId.llama3_1_8b_instruct.value,
+    build_model_alias(
+        "llama3.1:8b-instruct-fp16",
+        CoreModelId.llama3_1_8b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="llama3.1:70b-instruct-fp16",
-        aliases=["Llama3.1-70B-Instruct"],
-        llama_model=CoreModelId.llama3_1_70b_instruct.value,
+    build_model_alias(
+        "llama3.1:70b-instruct-fp16",
+        CoreModelId.llama3_1_70b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="llama3.2:1b-instruct-fp16",
-        aliases=["Llama3.2-1B-Instruct"],
-        llama_model=CoreModelId.llama3_2_1b_instruct.value,
+    build_model_alias(
+        "llama3.2:1b-instruct-fp16",
+        CoreModelId.llama3_2_1b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="llama3.2:3b-instruct-fp16",
-        aliases=["Llama3.2-3B-Instruct"],
-        llama_model=CoreModelId.llama3_2_3b_instruct.value,
+    build_model_alias(
+        "llama3.2:3b-instruct-fp16",
+        CoreModelId.llama3_2_3b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="llama-guard3:8b",
-        aliases=["Llama-Guard-3-8B"],
-        llama_model=CoreModelId.llama_guard_3_8b.value,
+    build_model_alias(
+        "llama-guard3:8b",
+        CoreModelId.llama_guard_3_8b,
     ),
-    ModelAlias(
-        provider_model_id="llama-guard3:1b",
-        aliases=["Llama-Guard-3-1B"],
-        llama_model=CoreModelId.llama_guard_3_1b.value,
+    build_model_alias(
+        "llama-guard3:1b",
+        CoreModelId.llama_guard_3_1b,
     ),
-    ModelAlias(
-        provider_model_id="x/llama3.2-vision:11b-instruct-fp16",
-        aliases=["Llama3.2-11B-Vision-Instruct"],
-        llama_model=CoreModelId.llama3_2_11b_vision_instruct.value,
+    build_model_alias(
+        "x/llama3.2-vision:11b-instruct-fp16",
+        CoreModelId.llama3_2_11b_vision_instruct,
     ),
 ]
 
diff --git a/llama_stack/providers/remote/inference/together/together.py b/llama_stack/providers/remote/inference/together/together.py
index 75f93f64f..644302a0f 100644
--- a/llama_stack/providers/remote/inference/together/together.py
+++ b/llama_stack/providers/remote/inference/together/together.py
@@ -18,7 +18,7 @@ from together import Together
 from llama_stack.apis.inference import *  # noqa: F403
 from llama_stack.distribution.request_headers import NeedsRequestProviderData
 from llama_stack.providers.utils.inference.model_registry import (
-    ModelAlias,
+    build_model_alias,
     ModelRegistryHelper,
 )
 from llama_stack.providers.utils.inference.openai_compat import (
@@ -39,45 +39,37 @@ from .config import TogetherImplConfig
 
 
 model_aliases = [
-    ModelAlias(
-        provider_model_id="meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo",
-        aliases=["Llama3.1-8B-Instruct"],
-        llama_model=CoreModelId.llama3_1_8b_instruct.value,
+    build_model_alias(
+        "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo",
+        CoreModelId.llama3_1_8b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo",
-        aliases=["Llama3.1-70B-Instruct"],
-        llama_model=CoreModelId.llama3_1_70b_instruct.value,
+    build_model_alias(
+        "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo",
+        CoreModelId.llama3_1_70b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo",
-        aliases=["Llama3.1-405B-Instruct"],
-        llama_model=CoreModelId.llama3_1_405b_instruct.value,
+    build_model_alias(
+        "meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo",
+        CoreModelId.llama3_1_405b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta-llama/Llama-3.2-3B-Instruct-Turbo",
-        aliases=["Llama3.2-3B-Instruct"],
-        llama_model=CoreModelId.llama3_2_3b_instruct.value,
+    build_model_alias(
+        "meta-llama/Llama-3.2-3B-Instruct-Turbo",
+        CoreModelId.llama3_2_3b_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo",
-        aliases=["Llama3.2-11B-Vision-Instruct"],
-        llama_model=CoreModelId.llama3_2_11b_vision_instruct.value,
+    build_model_alias(
+        "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo",
+        CoreModelId.llama3_2_11b_vision_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo",
-        aliases=["Llama3.2-90B-Vision-Instruct"],
-        llama_model=CoreModelId.llama3_2_90b_vision_instruct.value,
+    build_model_alias(
+        "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo",
+        CoreModelId.llama3_2_90b_vision_instruct,
     ),
-    ModelAlias(
-        provider_model_id="meta-llama/Meta-Llama-Guard-3-8B",
-        aliases=["Llama-Guard-3-8B"],
-        llama_model=CoreModelId.llama_guard_3_8b.value,
+    build_model_alias(
+        "meta-llama/Meta-Llama-Guard-3-8B",
+        CoreModelId.llama_guard_3_8b,
     ),
-    ModelAlias(
-        provider_model_id="meta-llama/Llama-Guard-3-11B-Vision-Turbo",
-        aliases=["Llama-Guard-3-11B-Vision"],
-        llama_model=CoreModelId.llama_guard_3_11b_vision.value,
+    build_model_alias(
+        "meta-llama/Llama-Guard-3-11B-Vision-Turbo",
+        CoreModelId.llama_guard_3_11b_vision,
     ),
 ]
 
diff --git a/llama_stack/providers/remote/inference/vllm/vllm.py b/llama_stack/providers/remote/inference/vllm/vllm.py
index c49541fd9..9bf25c5ad 100644
--- a/llama_stack/providers/remote/inference/vllm/vllm.py
+++ b/llama_stack/providers/remote/inference/vllm/vllm.py
@@ -16,7 +16,7 @@ from llama_stack.apis.inference import *  # noqa: F403
 from llama_stack.providers.datatypes import ModelsProtocolPrivate
 
 from llama_stack.providers.utils.inference.model_registry import (
-    ModelAlias,
+    build_model_alias,
     ModelRegistryHelper,
 )
 from llama_stack.providers.utils.inference.openai_compat import (
@@ -36,10 +36,9 @@ from .config import VLLMInferenceAdapterConfig
 
 def build_model_aliases():
     return [
-        ModelAlias(
-            provider_model_id=model.huggingface_repo,
-            aliases=[model.descriptor()],
-            llama_model=model.descriptor(),
+        build_model_alias(
+            model.huggingface_repo,
+            model.core_model_id,
         )
         for model in all_registered_models()
         if model.huggingface_repo
@@ -55,11 +54,6 @@ class VLLMInferenceAdapter(Inference, ModelRegistryHelper, ModelsProtocolPrivate
         self.config = config
         self.formatter = ChatFormat(Tokenizer.get_instance())
         self.client = None
-        self.huggingface_repo_to_llama_model_id = {
-            model.huggingface_repo: model.descriptor()
-            for model in all_registered_models()
-            if model.huggingface_repo
-        }
 
     async def initialize(self) -> None:
         self.client = OpenAI(base_url=self.config.url, api_key=self.config.api_token)
diff --git a/llama_stack/providers/utils/inference/model_registry.py b/llama_stack/providers/utils/inference/model_registry.py
index b3401d8f5..35d67a4cc 100644
--- a/llama_stack/providers/utils/inference/model_registry.py
+++ b/llama_stack/providers/utils/inference/model_registry.py
@@ -5,13 +5,35 @@
 # the root directory of this source tree.
 
 from collections import namedtuple
-from typing import List
+from typing import List, Optional
+
+from llama_models.datatypes import CoreModelId
+from llama_models.sku_list import all_registered_models
 
 from llama_stack.providers.datatypes import Model, ModelsProtocolPrivate
 
 ModelAlias = namedtuple("ModelAlias", ["provider_model_id", "aliases", "llama_model"])
 
 
+def get_huggingface_repo(core_model_id: CoreModelId) -> Optional[str]:
+    """Get the Hugging Face repository for a given CoreModelId."""
+    for model in all_registered_models():
+        if model.core_model_id == core_model_id:
+            return model.huggingface_repo
+    return None
+
+
+def build_model_alias(provider_model_id: str, core_model_id: CoreModelId) -> ModelAlias:
+    return ModelAlias(
+        provider_model_id=provider_model_id,
+        aliases=[
+            core_model_id.value,
+            get_huggingface_repo(core_model_id),
+        ],
+        llama_model=core_model_id.value,
+    )
+
+
 class ModelLookup:
     def __init__(
         self,