test: verification on provider's OAI endpoints (#1893)

# What does this PR do?


## Test Plan
export MODEL=accounts/fireworks/models/llama4-scout-instruct-basic;
LLAMA_STACK_CONFIG=verification pytest -s -v tests/integration/inference
--vision-model $MODEL --text-model $MODEL

docs/source/distributions/self_hosted_distro/fireworks.md

@@ -46,6 +46,8 @@ The following models are available by default:
 - `accounts/fireworks/models/llama-v3p3-70b-instruct (aliases: meta-llama/Llama-3.3-70B-Instruct)`
 - `accounts/fireworks/models/llama-guard-3-8b (aliases: meta-llama/Llama-Guard-3-8B)`
 - `accounts/fireworks/models/llama-guard-3-11b-vision (aliases: meta-llama/Llama-Guard-3-11B-Vision)`
+- `accounts/fireworks/models/llama4-scout-instruct-basic (aliases: meta-llama/Llama-4-Scout-17B-16E-Instruct)`
+- `accounts/fireworks/models/llama4-maverick-instruct-basic (aliases: meta-llama/Llama-4-Maverick-17B-128E-Instruct)`
 - `nomic-ai/nomic-embed-text-v1.5 `
 
 

docs/source/distributions/self_hosted_distro/groq.md

@@ -42,6 +42,8 @@ The following models are available by default:
 - `groq/llama3-70b-8192 (aliases: meta-llama/Llama-3-70B-Instruct)`
 - `groq/llama-3.3-70b-versatile (aliases: meta-llama/Llama-3.3-70B-Instruct)`
 - `groq/llama-3.2-3b-preview (aliases: meta-llama/Llama-3.2-3B-Instruct)`
+- `groq/llama-4-scout-17b-16e-instruct (aliases: meta-llama/Llama-4-Scout-17B-16E-Instruct)`
+- `groq/llama-4-maverick-17b-128e-instruct (aliases: meta-llama/Llama-4-Maverick-17B-128E-Instruct)`
 
 
 ### Prerequisite: API Keys

docs/source/distributions/self_hosted_distro/sambanova.md

@@ -43,6 +43,7 @@ The following models are available by default:
 - `Llama-3.2-11B-Vision-Instruct (aliases: meta-llama/Llama-3.2-11B-Vision-Instruct)`
 - `Llama-3.2-90B-Vision-Instruct (aliases: meta-llama/Llama-3.2-90B-Vision-Instruct)`
 - `Meta-Llama-Guard-3-8B (aliases: meta-llama/Llama-Guard-3-8B)`
+- `Llama-4-Scout-17B-16E-Instruct (aliases: meta-llama/Llama-4-Scout-17B-16E-Instruct)`
 
 
 ### Prerequisite: API Keys

docs/source/distributions/self_hosted_distro/together.md

@@ -48,6 +48,8 @@ The following models are available by default:
 - `meta-llama/Llama-Guard-3-11B-Vision-Turbo (aliases: meta-llama/Llama-Guard-3-11B-Vision)`
 - `togethercomputer/m2-bert-80M-8k-retrieval `
 - `togethercomputer/m2-bert-80M-32k-retrieval `
+- `meta-llama/Llama-4-Scout-17B-16E-Instruct (aliases: meta-llama/Llama-4-Scout-17B-16E-Instruct, together/meta-llama/Llama-4-Scout-17B-16E-Instruct)`
+- `meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 (aliases: meta-llama/Llama-4-Maverick-17B-128E-Instruct, together/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8)`
 
 
 ### Prerequisite: API Keys

llama_stack/providers/registry/inference.py

@@ -217,6 +217,56 @@ def available_providers() -> List[ProviderSpec]:
                 provider_data_validator="llama_stack.providers.remote.inference.groq.config.GroqProviderDataValidator",
             ),
         ),
+        remote_provider_spec(
+            api=Api.inference,
+            adapter=AdapterSpec(
+                adapter_type="fireworks-openai-compat",
+                pip_packages=["litellm"],
+                module="llama_stack.providers.remote.inference.fireworks_openai_compat",
+                config_class="llama_stack.providers.remote.inference.fireworks_openai_compat.config.FireworksCompatConfig",
+                provider_data_validator="llama_stack.providers.remote.inference.fireworks_openai_compat.config.FireworksProviderDataValidator",
+            ),
+        ),
+        remote_provider_spec(
+            api=Api.inference,
+            adapter=AdapterSpec(
+                adapter_type="together-openai-compat",
+                pip_packages=["litellm"],
+                module="llama_stack.providers.remote.inference.together_openai_compat",
+                config_class="llama_stack.providers.remote.inference.together_openai_compat.config.TogetherCompatConfig",
+                provider_data_validator="llama_stack.providers.remote.inference.together_openai_compat.config.TogetherProviderDataValidator",
+            ),
+        ),
+        remote_provider_spec(
+            api=Api.inference,
+            adapter=AdapterSpec(
+                adapter_type="groq-openai-compat",
+                pip_packages=["litellm"],
+                module="llama_stack.providers.remote.inference.groq_openai_compat",
+                config_class="llama_stack.providers.remote.inference.groq_openai_compat.config.GroqCompatConfig",
+                provider_data_validator="llama_stack.providers.remote.inference.groq_openai_compat.config.GroqProviderDataValidator",
+            ),
+        ),
+        remote_provider_spec(
+            api=Api.inference,
+            adapter=AdapterSpec(
+                adapter_type="sambanova-openai-compat",
+                pip_packages=["litellm"],
+                module="llama_stack.providers.remote.inference.sambanova_openai_compat",
+                config_class="llama_stack.providers.remote.inference.sambanova_openai_compat.config.SambaNovaCompatConfig",
+                provider_data_validator="llama_stack.providers.remote.inference.sambanova_openai_compat.config.SambaNovaProviderDataValidator",
+            ),
+        ),
+        remote_provider_spec(
+            api=Api.inference,
+            adapter=AdapterSpec(
+                adapter_type="cerebras-openai-compat",
+                pip_packages=["litellm"],
+                module="llama_stack.providers.remote.inference.cerebras_openai_compat",
+                config_class="llama_stack.providers.remote.inference.cerebras_openai_compat.config.CerebrasCompatConfig",
+                provider_data_validator="llama_stack.providers.remote.inference.cerebras_openai_compat.config.CerebrasProviderDataValidator",
+            ),
+        ),
         remote_provider_spec(
             api=Api.inference,
             adapter=AdapterSpec(

llama_stack/providers/remote/inference/cerebras_openai_compat/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.apis.inference import Inference
+
+from .config import CerebrasCompatConfig
+
+
+async def get_adapter_impl(config: CerebrasCompatConfig, _deps) -> Inference:
+    # import dynamically so the import is used only when it is needed
+    from .cerebras import CerebrasCompatInferenceAdapter
+
+    adapter = CerebrasCompatInferenceAdapter(config)
+    return adapter

llama_stack/providers/remote/inference/cerebras_openai_compat/cerebras.py

@@ -0,0 +1,30 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.providers.remote.inference.cerebras_openai_compat.config import CerebrasCompatConfig
+from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
+
+from ..cerebras.models import MODEL_ENTRIES
+
+
+class CerebrasCompatInferenceAdapter(LiteLLMOpenAIMixin):
+    _config: CerebrasCompatConfig
+
+    def __init__(self, config: CerebrasCompatConfig):
+        LiteLLMOpenAIMixin.__init__(
+            self,
+            model_entries=MODEL_ENTRIES,
+            api_key_from_config=config.api_key,
+            provider_data_api_key_field="cerebras_api_key",
+            openai_compat_api_base=config.openai_compat_api_base,
+        )
+        self.config = config
+
+    async def initialize(self):
+        await super().initialize()
+
+    async def shutdown(self):
+        await super().shutdown()

llama_stack/providers/remote/inference/cerebras_openai_compat/config.py

@@ -0,0 +1,38 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import Any, Dict, Optional
+
+from pydantic import BaseModel, Field
+
+from llama_stack.schema_utils import json_schema_type
+
+
+class CerebrasProviderDataValidator(BaseModel):
+    cerebras_api_key: Optional[str] = Field(
+        default=None,
+        description="API key for Cerebras models",
+    )
+
+
+@json_schema_type
+class CerebrasCompatConfig(BaseModel):
+    api_key: Optional[str] = Field(
+        default=None,
+        description="The Cerebras API key",
+    )
+
+    openai_compat_api_base: str = Field(
+        default="https://api.cerebras.ai/v1",
+        description="The URL for the Cerebras API server",
+    )
+
+    @classmethod
+    def sample_run_config(cls, api_key: str = "${env.CEREBRAS_API_KEY}", **kwargs) -> Dict[str, Any]:
+        return {
+            "openai_compat_api_base": "https://api.cerebras.ai/v1",
+            "api_key": api_key,
+        }

llama_stack/providers/remote/inference/fireworks/models.py

@@ -48,6 +48,14 @@ MODEL_ENTRIES = [
         "accounts/fireworks/models/llama-guard-3-11b-vision",
         CoreModelId.llama_guard_3_11b_vision.value,
     ),
+    build_hf_repo_model_entry(
+        "accounts/fireworks/models/llama4-scout-instruct-basic",
+        CoreModelId.llama4_scout_17b_16e_instruct.value,
+    ),
+    build_hf_repo_model_entry(
+        "accounts/fireworks/models/llama4-maverick-instruct-basic",
+        CoreModelId.llama4_maverick_17b_128e_instruct.value,
+    ),
     ProviderModelEntry(
         provider_model_id="nomic-ai/nomic-embed-text-v1.5",
         model_type=ModelType.embedding,

llama_stack/providers/remote/inference/fireworks_openai_compat/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.apis.inference import Inference
+
+from .config import FireworksCompatConfig
+
+
+async def get_adapter_impl(config: FireworksCompatConfig, _deps) -> Inference:
+    # import dynamically so the import is used only when it is needed
+    from .fireworks import FireworksCompatInferenceAdapter
+
+    adapter = FireworksCompatInferenceAdapter(config)
+    return adapter

llama_stack/providers/remote/inference/fireworks_openai_compat/config.py

@@ -0,0 +1,38 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import Any, Dict, Optional
+
+from pydantic import BaseModel, Field
+
+from llama_stack.schema_utils import json_schema_type
+
+
+class FireworksProviderDataValidator(BaseModel):
+    fireworks_api_key: Optional[str] = Field(
+        default=None,
+        description="API key for Fireworks models",
+    )
+
+
+@json_schema_type
+class FireworksCompatConfig(BaseModel):
+    api_key: Optional[str] = Field(
+        default=None,
+        description="The Fireworks API key",
+    )
+
+    openai_compat_api_base: str = Field(
+        default="https://api.fireworks.ai/inference/v1",
+        description="The URL for the Fireworks API server",
+    )
+
+    @classmethod
+    def sample_run_config(cls, api_key: str = "${env.FIREWORKS_API_KEY}", **kwargs) -> Dict[str, Any]:
+        return {
+            "openai_compat_api_base": "https://api.fireworks.ai/inference/v1",
+            "api_key": api_key,
+        }

llama_stack/providers/remote/inference/fireworks_openai_compat/fireworks.py

@@ -0,0 +1,30 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.providers.remote.inference.fireworks_openai_compat.config import FireworksCompatConfig
+from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
+
+from ..fireworks.models import MODEL_ENTRIES
+
+
+class FireworksCompatInferenceAdapter(LiteLLMOpenAIMixin):
+    _config: FireworksCompatConfig
+
+    def __init__(self, config: FireworksCompatConfig):
+        LiteLLMOpenAIMixin.__init__(
+            self,
+            model_entries=MODEL_ENTRIES,
+            api_key_from_config=config.api_key,
+            provider_data_api_key_field="fireworks_api_key",
+            openai_compat_api_base=config.openai_compat_api_base,
+        )
+        self.config = config
+
+    async def initialize(self):
+        await super().initialize()
+
+    async def shutdown(self):
+        await super().shutdown()

llama_stack/providers/remote/inference/groq/models.py

@@ -35,4 +35,12 @@ MODEL_ENTRIES = [
         "groq/llama-3.2-3b-preview",
         CoreModelId.llama3_2_3b_instruct.value,
     ),
+    build_hf_repo_model_entry(
+        "groq/llama-4-scout-17b-16e-instruct",
+        CoreModelId.llama4_scout_17b_16e_instruct.value,
+    ),
+    build_hf_repo_model_entry(
+        "groq/llama-4-maverick-17b-128e-instruct",
+        CoreModelId.llama4_maverick_17b_128e_instruct.value,
+    ),
 ]

llama_stack/providers/remote/inference/groq_openai_compat/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.apis.inference import Inference
+
+from .config import GroqCompatConfig
+
+
+async def get_adapter_impl(config: GroqCompatConfig, _deps) -> Inference:
+    # import dynamically so the import is used only when it is needed
+    from .groq import GroqCompatInferenceAdapter
+
+    adapter = GroqCompatInferenceAdapter(config)
+    return adapter

llama_stack/providers/remote/inference/groq_openai_compat/config.py

@@ -0,0 +1,38 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import Any, Dict, Optional
+
+from pydantic import BaseModel, Field
+
+from llama_stack.schema_utils import json_schema_type
+
+
+class GroqProviderDataValidator(BaseModel):
+    groq_api_key: Optional[str] = Field(
+        default=None,
+        description="API key for Groq models",
+    )
+
+
+@json_schema_type
+class GroqCompatConfig(BaseModel):
+    api_key: Optional[str] = Field(
+        default=None,
+        description="The Groq API key",
+    )
+
+    openai_compat_api_base: str = Field(
+        default="https://api.groq.com/openai/v1",
+        description="The URL for the Groq API server",
+    )
+
+    @classmethod
+    def sample_run_config(cls, api_key: str = "${env.GROQ_API_KEY}", **kwargs) -> Dict[str, Any]:
+        return {
+            "openai_compat_api_base": "https://api.groq.com/openai/v1",
+            "api_key": api_key,
+        }

llama_stack/providers/remote/inference/groq_openai_compat/groq.py

@@ -0,0 +1,30 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.providers.remote.inference.groq_openai_compat.config import GroqCompatConfig
+from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
+
+from ..groq.models import MODEL_ENTRIES
+
+
+class GroqCompatInferenceAdapter(LiteLLMOpenAIMixin):
+    _config: GroqCompatConfig
+
+    def __init__(self, config: GroqCompatConfig):
+        LiteLLMOpenAIMixin.__init__(
+            self,
+            model_entries=MODEL_ENTRIES,
+            api_key_from_config=config.api_key,
+            provider_data_api_key_field="groq_api_key",
+            openai_compat_api_base=config.openai_compat_api_base,
+        )
+        self.config = config
+
+    async def initialize(self):
+        await super().initialize()
+
+    async def shutdown(self):
+        await super().shutdown()

llama_stack/providers/remote/inference/sambanova/models.py

@@ -46,4 +46,8 @@ MODEL_ENTRIES = [
         "Meta-Llama-Guard-3-8B",
         CoreModelId.llama_guard_3_8b.value,
     ),
+    build_hf_repo_model_entry(
+        "Llama-4-Scout-17B-16E-Instruct",
+        CoreModelId.llama4_scout_17b_16e_instruct.value,
+    ),
 ]

llama_stack/providers/remote/inference/sambanova_openai_compat/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.apis.inference import Inference
+
+from .config import SambaNovaCompatConfig
+
+
+async def get_adapter_impl(config: SambaNovaCompatConfig, _deps) -> Inference:
+    # import dynamically so the import is used only when it is needed
+    from .sambanova import SambaNovaCompatInferenceAdapter
+
+    adapter = SambaNovaCompatInferenceAdapter(config)
+    return adapter

llama_stack/providers/remote/inference/sambanova_openai_compat/config.py

@@ -0,0 +1,38 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import Any, Dict, Optional
+
+from pydantic import BaseModel, Field
+
+from llama_stack.schema_utils import json_schema_type
+
+
+class SambaNovaProviderDataValidator(BaseModel):
+    sambanova_api_key: Optional[str] = Field(
+        default=None,
+        description="API key for SambaNova models",
+    )
+
+
+@json_schema_type
+class SambaNovaCompatConfig(BaseModel):
+    api_key: Optional[str] = Field(
+        default=None,
+        description="The SambaNova API key",
+    )
+
+    openai_compat_api_base: str = Field(
+        default="https://api.sambanova.ai/v1",
+        description="The URL for the SambaNova API server",
+    )
+
+    @classmethod
+    def sample_run_config(cls, api_key: str = "${env.SAMBANOVA_API_KEY}", **kwargs) -> Dict[str, Any]:
+        return {
+            "openai_compat_api_base": "https://api.sambanova.ai/v1",
+            "api_key": api_key,
+        }

llama_stack/providers/remote/inference/sambanova_openai_compat/sambanova.py

@@ -0,0 +1,30 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.providers.remote.inference.sambanova_openai_compat.config import SambaNovaCompatConfig
+from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
+
+from ..sambanova.models import MODEL_ENTRIES
+
+
+class SambaNovaCompatInferenceAdapter(LiteLLMOpenAIMixin):
+    _config: SambaNovaCompatConfig
+
+    def __init__(self, config: SambaNovaCompatConfig):
+        LiteLLMOpenAIMixin.__init__(
+            self,
+            model_entries=MODEL_ENTRIES,
+            api_key_from_config=config.api_key,
+            provider_data_api_key_field="sambanova_api_key",
+            openai_compat_api_base=config.openai_compat_api_base,
+        )
+        self.config = config
+
+    async def initialize(self):
+        await super().initialize()
+
+    async def shutdown(self):
+        await super().shutdown()

llama_stack/providers/remote/inference/together/models.py

@@ -64,4 +64,18 @@ MODEL_ENTRIES = [
             "context_length": 32768,
         },
     ),
+    build_hf_repo_model_entry(
+        "meta-llama/Llama-4-Scout-17B-16E-Instruct",
+        CoreModelId.llama4_scout_17b_16e_instruct.value,
+        additional_aliases=[
+            "together/meta-llama/Llama-4-Scout-17B-16E-Instruct",
+        ],
+    ),
+    build_hf_repo_model_entry(
+        "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
+        CoreModelId.llama4_maverick_17b_128e_instruct.value,
+        additional_aliases=[
+            "together/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
+        ],
+    ),
 ]

llama_stack/providers/remote/inference/together/together.py

@@ -118,7 +118,7 @@ class TogetherInferenceAdapter(ModelRegistryHelper, Inference, NeedsRequestProvi
 
     async def _stream_completion(self, request: CompletionRequest) -> AsyncGenerator:
         params = await self._get_params(request)
-        client = await self._get_client()
+        client = self._get_client()
         stream = await client.completions.create(**params)
         async for chunk in process_completion_stream_response(stream):
             yield chunk

llama_stack/providers/remote/inference/together_openai_compat/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.apis.inference import Inference
+
+from .config import TogetherCompatConfig
+
+
+async def get_adapter_impl(config: TogetherCompatConfig, _deps) -> Inference:
+    # import dynamically so the import is used only when it is needed
+    from .together import TogetherCompatInferenceAdapter
+
+    adapter = TogetherCompatInferenceAdapter(config)
+    return adapter

llama_stack/providers/remote/inference/together_openai_compat/config.py

@@ -0,0 +1,38 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import Any, Dict, Optional
+
+from pydantic import BaseModel, Field
+
+from llama_stack.schema_utils import json_schema_type
+
+
+class TogetherProviderDataValidator(BaseModel):
+    together_api_key: Optional[str] = Field(
+        default=None,
+        description="API key for Together models",
+    )
+
+
+@json_schema_type
+class TogetherCompatConfig(BaseModel):
+    api_key: Optional[str] = Field(
+        default=None,
+        description="The Together API key",
+    )
+
+    openai_compat_api_base: str = Field(
+        default="https://api.together.xyz/v1",
+        description="The URL for the Together API server",
+    )
+
+    @classmethod
+    def sample_run_config(cls, api_key: str = "${env.TOGETHER_API_KEY}", **kwargs) -> Dict[str, Any]:
+        return {
+            "openai_compat_api_base": "https://api.together.xyz/v1",
+            "api_key": api_key,
+        }

llama_stack/providers/remote/inference/together_openai_compat/together.py

@@ -0,0 +1,30 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from llama_stack.providers.remote.inference.together_openai_compat.config import TogetherCompatConfig
+from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
+
+from ..together.models import MODEL_ENTRIES
+
+
+class TogetherCompatInferenceAdapter(LiteLLMOpenAIMixin):
+    _config: TogetherCompatConfig
+
+    def __init__(self, config: TogetherCompatConfig):
+        LiteLLMOpenAIMixin.__init__(
+            self,
+            model_entries=MODEL_ENTRIES,
+            api_key_from_config=config.api_key,
+            provider_data_api_key_field="together_api_key",
+            openai_compat_api_base=config.openai_compat_api_base,
+        )
+        self.config = config
+
+    async def initialize(self):
+        await super().initialize()
+
+    async def shutdown(self):
+        await super().shutdown()

llama_stack/providers/utils/inference/model_registry.py

@@ -33,12 +33,17 @@ def get_huggingface_repo(model_descriptor: str) -> Optional[str]:
     return None
 
 
-def build_hf_repo_model_entry(provider_model_id: str, model_descriptor: str) -> ProviderModelEntry:
+def build_hf_repo_model_entry(
+    provider_model_id: str, model_descriptor: str, additional_aliases: Optional[List[str]] = None
+) -> ProviderModelEntry:
+    aliases = [
+        get_huggingface_repo(model_descriptor),
+    ]
+    if additional_aliases:
+        aliases.extend(additional_aliases)
     return ProviderModelEntry(
         provider_model_id=provider_model_id,
-        aliases=[
-            get_huggingface_repo(model_descriptor),
-        ],
+        aliases=aliases,
         llama_model=model_descriptor,
     )
 

llama_stack/providers/utils/inference/openai_compat.py

@@ -642,6 +642,36 @@ PYTHON_TYPE_TO_LITELLM_TYPE = {
 }
 
 
+def to_openai_param_type(param_type: str) -> dict:
+    """
+    Convert Python type hints to OpenAI parameter type format.
+
+    Examples:
+        'str' -> {'type': 'string'}
+        'int' -> {'type': 'integer'}
+        'list[str]' -> {'type': 'array', 'items': {'type': 'string'}}
+        'list[int]' -> {'type': 'array', 'items': {'type': 'integer'}}
+    """
+    # Handle basic types first
+    basic_types = {
+        "str": "string",
+        "int": "integer",
+        "float": "number",
+        "bool": "boolean",
+    }
+
+    if param_type in basic_types:
+        return {"type": basic_types[param_type]}
+
+    # Handle list/array types
+    if param_type.startswith("list[") and param_type.endswith("]"):
+        inner_type = param_type[5:-1]
+        if inner_type in basic_types:
+            return {"type": "array", "items": {"type": basic_types.get(inner_type, inner_type)}}
+
+    return {"type": param_type}
+
+
 def convert_tooldef_to_openai_tool(tool: ToolDefinition) -> dict:
     """
     Convert a ToolDefinition to an OpenAI API-compatible dictionary.
@@ -702,7 +732,7 @@ def convert_tooldef_to_openai_tool(tool: ToolDefinition) -> dict:
         properties = parameters["properties"]
         required = []
         for param_name, param in tool.parameters.items():
-            properties[param_name] = {"type": PYTHON_TYPE_TO_LITELLM_TYPE.get(param.param_type, param.param_type)}
+            properties[param_name] = to_openai_param_type(param.param_type)
             if param.description:
                 properties[param_name].update(description=param.description)
             if param.default:

llama_stack/templates/ci-tests/run.yaml

@@ -193,6 +193,26 @@ models:
   provider_id: fireworks
   provider_model_id: accounts/fireworks/models/llama-guard-3-11b-vision
   model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
 - metadata:
     embedding_dimension: 768
     context_length: 8192

llama_stack/templates/dependencies.json

@@ -682,6 +682,45 @@
     "sentence-transformers --no-deps",
     "torch torchvision --index-url https://download.pytorch.org/whl/cpu"
   ],
+  "verification": [
+    "aiosqlite",
+    "autoevals",
+    "blobfile",
+    "chardet",
+    "chromadb-client",
+    "datasets",
+    "emoji",
+    "fastapi",
+    "fire",
+    "httpx",
+    "langdetect",
+    "litellm",
+    "matplotlib",
+    "mcp",
+    "nltk",
+    "numpy",
+    "openai",
+    "opentelemetry-exporter-otlp-proto-http",
+    "opentelemetry-sdk",
+    "pandas",
+    "pillow",
+    "psycopg2-binary",
+    "pymongo",
+    "pypdf",
+    "pythainlp",
+    "redis",
+    "requests",
+    "scikit-learn",
+    "scipy",
+    "sentencepiece",
+    "sqlite-vec",
+    "tqdm",
+    "transformers",
+    "tree_sitter",
+    "uvicorn",
+    "sentence-transformers --no-deps",
+    "torch torchvision --index-url https://download.pytorch.org/whl/cpu"
+  ],
   "vllm-gpu": [
     "aiosqlite",
     "autoevals",

llama_stack/templates/dev/run.yaml

@@ -251,6 +251,26 @@ models:
   provider_id: fireworks
   provider_model_id: accounts/fireworks/models/llama-guard-3-11b-vision
   model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
 - metadata:
     embedding_dimension: 768
     context_length: 8192
@@ -356,6 +376,26 @@ models:
   provider_id: groq
   provider_model_id: groq/llama-3.2-3b-preview
   model_type: llm
+- metadata: {}
+  model_id: groq/llama-4-scout-17b-16e-instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-scout-17b-16e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-scout-17b-16e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama-4-maverick-17b-128e-instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-maverick-17b-128e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-maverick-17b-128e-instruct
+  model_type: llm
 - metadata:
     embedding_dimension: 384
   model_id: all-MiniLM-L6-v2

llama_stack/templates/fireworks/run-with-safety.yaml

@@ -205,6 +205,26 @@ models:
   provider_id: fireworks
   provider_model_id: accounts/fireworks/models/llama-guard-3-11b-vision
   model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
 - metadata:
     embedding_dimension: 768
     context_length: 8192

llama_stack/templates/fireworks/run.yaml

@@ -200,6 +200,26 @@ models:
   provider_id: fireworks
   provider_model_id: accounts/fireworks/models/llama-guard-3-11b-vision
   model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: fireworks
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
 - metadata:
     embedding_dimension: 768
     context_length: 8192

llama_stack/templates/groq/run.yaml

@@ -148,6 +148,26 @@ models:
   provider_id: groq
   provider_model_id: groq/llama-3.2-3b-preview
   model_type: llm
+- metadata: {}
+  model_id: groq/llama-4-scout-17b-16e-instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-scout-17b-16e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-scout-17b-16e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama-4-maverick-17b-128e-instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-maverick-17b-128e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: groq
+  provider_model_id: groq/llama-4-maverick-17b-128e-instruct
+  model_type: llm
 - metadata:
     embedding_dimension: 384
   model_id: all-MiniLM-L6-v2

llama_stack/templates/sambanova/run.yaml

@@ -165,6 +165,16 @@ models:
   provider_id: sambanova
   provider_model_id: Meta-Llama-Guard-3-8B
   model_type: llm
+- metadata: {}
+  model_id: Llama-4-Scout-17B-16E-Instruct
+  provider_id: sambanova
+  provider_model_id: Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: sambanova
+  provider_model_id: Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
 shields:
 - shield_id: meta-llama/Llama-Guard-3-8B
 vector_dbs: []

llama_stack/templates/together/run-with-safety.yaml

@@ -219,6 +219,36 @@ models:
   provider_id: together
   provider_model_id: togethercomputer/m2-bert-80M-32k-retrieval
   model_type: embedding
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: together/meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
+- metadata: {}
+  model_id: together/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
 - metadata:
     embedding_dimension: 384
   model_id: all-MiniLM-L6-v2

llama_stack/templates/together/run.yaml

@@ -214,6 +214,36 @@ models:
   provider_id: together
   provider_model_id: togethercomputer/m2-bert-80M-32k-retrieval
   model_type: embedding
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: together/meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
+- metadata: {}
+  model_id: together/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  provider_id: together
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
 - metadata:
     embedding_dimension: 384
   model_id: all-MiniLM-L6-v2

llama_stack/templates/verification/__init__.py

@@ -0,0 +1,7 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from .verification import get_distribution_template  # noqa: F401

llama_stack/templates/verification/build.yaml

@@ -0,0 +1,38 @@
+version: '2'
+distribution_spec:
+  description: Distribution for running e2e tests in CI
+  providers:
+    inference:
+    - remote::openai
+    - remote::fireworks-openai-compat
+    - remote::together-openai-compat
+    - remote::groq-openai-compat
+    - remote::sambanova-openai-compat
+    - remote::cerebras-openai-compat
+    - inline::sentence-transformers
+    vector_io:
+    - inline::sqlite-vec
+    - remote::chromadb
+    - remote::pgvector
+    safety:
+    - inline::llama-guard
+    agents:
+    - inline::meta-reference
+    telemetry:
+    - inline::meta-reference
+    eval:
+    - inline::meta-reference
+    datasetio:
+    - remote::huggingface
+    - inline::localfs
+    scoring:
+    - inline::basic
+    - inline::llm-as-judge
+    - inline::braintrust
+    tool_runtime:
+    - remote::brave-search
+    - remote::tavily-search
+    - inline::code-interpreter
+    - inline::rag-runtime
+    - remote::model-context-protocol
+image_type: conda

llama_stack/templates/verification/run.yaml

@@ -0,0 +1,626 @@
+version: '2'
+image_name: verification
+apis:
+- agents
+- datasetio
+- eval
+- inference
+- safety
+- scoring
+- telemetry
+- tool_runtime
+- vector_io
+providers:
+  inference:
+  - provider_id: openai
+    provider_type: remote::openai
+    config:
+      api_key: ${env.OPENAI_API_KEY:}
+  - provider_id: fireworks-openai-compat
+    provider_type: remote::fireworks-openai-compat
+    config:
+      openai_compat_api_base: https://api.fireworks.ai/inference/v1
+      api_key: ${env.FIREWORKS_API_KEY:}
+  - provider_id: together-openai-compat
+    provider_type: remote::together-openai-compat
+    config:
+      openai_compat_api_base: https://api.together.xyz/v1
+      api_key: ${env.TOGETHER_API_KEY:}
+  - provider_id: groq-openai-compat
+    provider_type: remote::groq-openai-compat
+    config:
+      openai_compat_api_base: https://api.groq.com/openai/v1
+      api_key: ${env.GROQ_API_KEY:}
+  - provider_id: sambanova-openai-compat
+    provider_type: remote::sambanova-openai-compat
+    config:
+      openai_compat_api_base: https://api.sambanova.ai/v1
+      api_key: ${env.SAMBANOVA_API_KEY:}
+  - provider_id: cerebras-openai-compat
+    provider_type: remote::cerebras-openai-compat
+    config:
+      openai_compat_api_base: https://api.cerebras.ai/v1
+      api_key: ${env.CEREBRAS_API_KEY:}
+  - provider_id: sentence-transformers
+    provider_type: inline::sentence-transformers
+    config: {}
+  vector_io:
+  - provider_id: sqlite-vec
+    provider_type: inline::sqlite-vec
+    config:
+      db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/verification}/sqlite_vec.db
+  - provider_id: ${env.ENABLE_CHROMADB+chromadb}
+    provider_type: remote::chromadb
+    config:
+      url: ${env.CHROMADB_URL:}
+  - provider_id: ${env.ENABLE_PGVECTOR+pgvector}
+    provider_type: remote::pgvector
+    config:
+      host: ${env.PGVECTOR_HOST:localhost}
+      port: ${env.PGVECTOR_PORT:5432}
+      db: ${env.PGVECTOR_DB:}
+      user: ${env.PGVECTOR_USER:}
+      password: ${env.PGVECTOR_PASSWORD:}
+  safety:
+  - provider_id: llama-guard
+    provider_type: inline::llama-guard
+    config:
+      excluded_categories: []
+  agents:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      persistence_store:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/verification}/agents_store.db
+  telemetry:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      service_name: "${env.OTEL_SERVICE_NAME:\u200B}"
+      sinks: ${env.TELEMETRY_SINKS:console,sqlite}
+      sqlite_db_path: ${env.SQLITE_DB_PATH:~/.llama/distributions/verification/trace_store.db}
+  eval:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/verification}/meta_reference_eval.db
+  datasetio:
+  - provider_id: huggingface
+    provider_type: remote::huggingface
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/verification}/huggingface_datasetio.db
+  - provider_id: localfs
+    provider_type: inline::localfs
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/verification}/localfs_datasetio.db
+  scoring:
+  - provider_id: basic
+    provider_type: inline::basic
+    config: {}
+  - provider_id: llm-as-judge
+    provider_type: inline::llm-as-judge
+    config: {}
+  - provider_id: braintrust
+    provider_type: inline::braintrust
+    config:
+      openai_api_key: ${env.OPENAI_API_KEY:}
+  tool_runtime:
+  - provider_id: brave-search
+    provider_type: remote::brave-search
+    config:
+      api_key: ${env.BRAVE_SEARCH_API_KEY:}
+      max_results: 3
+  - provider_id: tavily-search
+    provider_type: remote::tavily-search
+    config:
+      api_key: ${env.TAVILY_SEARCH_API_KEY:}
+      max_results: 3
+  - provider_id: code-interpreter
+    provider_type: inline::code-interpreter
+    config: {}
+  - provider_id: rag-runtime
+    provider_type: inline::rag-runtime
+    config: {}
+  - provider_id: model-context-protocol
+    provider_type: remote::model-context-protocol
+    config: {}
+metadata_store:
+  type: sqlite
+  db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/verification}/registry.db
+models:
+- metadata: {}
+  model_id: openai/gpt-4o
+  provider_id: openai
+  provider_model_id: openai/gpt-4o
+  model_type: llm
+- metadata: {}
+  model_id: openai/gpt-4o-mini
+  provider_id: openai
+  provider_model_id: openai/gpt-4o-mini
+  model_type: llm
+- metadata: {}
+  model_id: openai/chatgpt-4o-latest
+  provider_id: openai
+  provider_model_id: openai/chatgpt-4o-latest
+  model_type: llm
+- metadata:
+    embedding_dimension: 1536
+    context_length: 8192
+  model_id: openai/text-embedding-3-small
+  provider_id: openai
+  provider_model_id: openai/text-embedding-3-small
+  model_type: embedding
+- metadata:
+    embedding_dimension: 3072
+    context_length: 8192
+  model_id: openai/text-embedding-3-large
+  provider_id: openai
+  provider_model_id: openai/text-embedding-3-large
+  model_type: embedding
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-v3p1-8b-instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p1-8b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-8B-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p1-8b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-v3p1-70b-instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p1-70b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-70B-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p1-70b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-v3p1-405b-instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p1-405b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-405B-Instruct-FP8
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p1-405b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-v3p2-3b-instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p2-3b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-3B-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p2-3b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-v3p2-11b-vision-instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p2-11b-vision-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-11B-Vision-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p2-11b-vision-instruct
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-v3p2-90b-vision-instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p2-90b-vision-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-90B-Vision-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p2-90b-vision-instruct
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-v3p3-70b-instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p3-70b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.3-70B-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-v3p3-70b-instruct
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-guard-3-8b
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-guard-3-8b
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-Guard-3-8B
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-guard-3-8b
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama-guard-3-11b-vision
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-guard-3-11b-vision
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-Guard-3-11B-Vision
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama-guard-3-11b-vision
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama4-scout-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: fireworks-openai-compat
+  provider_model_id: accounts/fireworks/models/llama4-maverick-instruct-basic
+  model_type: llm
+- metadata:
+    embedding_dimension: 768
+    context_length: 8192
+  model_id: nomic-ai/nomic-embed-text-v1.5
+  provider_id: fireworks-openai-compat
+  provider_model_id: nomic-ai/nomic-embed-text-v1.5
+  model_type: embedding
+- metadata: {}
+  model_id: meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-8B-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-70B-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-405B-Instruct-FP8
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-3B-Instruct-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.2-3B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-3B-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.2-3B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-11B-Vision-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-90B-Vision-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.3-70B-Instruct-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.3-70B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.3-70B-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-3.3-70B-Instruct-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Meta-Llama-Guard-3-8B
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-Guard-3-8B
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-Guard-3-8B
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Meta-Llama-Guard-3-8B
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-Guard-3-11B-Vision-Turbo
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-Guard-3-11B-Vision-Turbo
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-Guard-3-11B-Vision
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-Guard-3-11B-Vision-Turbo
+  model_type: llm
+- metadata:
+    embedding_dimension: 768
+    context_length: 8192
+  model_id: togethercomputer/m2-bert-80M-8k-retrieval
+  provider_id: together-openai-compat
+  provider_model_id: togethercomputer/m2-bert-80M-8k-retrieval
+  model_type: embedding
+- metadata:
+    embedding_dimension: 768
+    context_length: 32768
+  model_id: togethercomputer/m2-bert-80M-32k-retrieval
+  provider_id: together-openai-compat
+  provider_model_id: togethercomputer/m2-bert-80M-32k-retrieval
+  model_type: embedding
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: together/meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
+- metadata: {}
+  model_id: together/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  provider_id: together-openai-compat
+  provider_model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama3-8b-8192
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama3-8b-8192
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-8B-Instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama3-8b-8192
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama-3.1-8b-instant
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-3.1-8b-instant
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama3-70b-8192
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama3-70b-8192
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3-70B-Instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama3-70b-8192
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama-3.3-70b-versatile
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-3.3-70b-versatile
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.3-70B-Instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-3.3-70b-versatile
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama-3.2-3b-preview
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-3.2-3b-preview
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-3B-Instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-3.2-3b-preview
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama-4-scout-17b-16e-instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-4-scout-17b-16e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-4-scout-17b-16e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: groq/llama-4-maverick-17b-128e-instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-4-maverick-17b-128e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Maverick-17B-128E-Instruct
+  provider_id: groq-openai-compat
+  provider_model_id: groq/llama-4-maverick-17b-128e-instruct
+  model_type: llm
+- metadata: {}
+  model_id: Meta-Llama-3.1-8B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.1-8B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-8B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.1-8B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Meta-Llama-3.1-70B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.1-70B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-70B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.1-70B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Meta-Llama-3.1-405B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.1-405B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-405B-Instruct-FP8
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.1-405B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Meta-Llama-3.2-1B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.2-1B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-1B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.2-1B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Meta-Llama-3.2-3B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.2-3B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-3B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.2-3B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Meta-Llama-3.3-70B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.3-70B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.3-70B-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-3.3-70B-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Llama-3.2-11B-Vision-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Llama-3.2-11B-Vision-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-11B-Vision-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Llama-3.2-11B-Vision-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Llama-3.2-90B-Vision-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Llama-3.2-90B-Vision-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.2-90B-Vision-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Llama-3.2-90B-Vision-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: Meta-Llama-Guard-3-8B
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-Guard-3-8B
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-Guard-3-8B
+  provider_id: sambanova-openai-compat
+  provider_model_id: Meta-Llama-Guard-3-8B
+  model_type: llm
+- metadata: {}
+  model_id: Llama-4-Scout-17B-16E-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-4-Scout-17B-16E-Instruct
+  provider_id: sambanova-openai-compat
+  provider_model_id: Llama-4-Scout-17B-16E-Instruct
+  model_type: llm
+- metadata: {}
+  model_id: llama3.1-8b
+  provider_id: cerebras-openai-compat
+  provider_model_id: llama3.1-8b
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.1-8B-Instruct
+  provider_id: cerebras-openai-compat
+  provider_model_id: llama3.1-8b
+  model_type: llm
+- metadata: {}
+  model_id: llama-3.3-70b
+  provider_id: cerebras-openai-compat
+  provider_model_id: llama-3.3-70b
+  model_type: llm
+- metadata: {}
+  model_id: meta-llama/Llama-3.3-70B-Instruct
+  provider_id: cerebras-openai-compat
+  provider_model_id: llama-3.3-70b
+  model_type: llm
+- metadata:
+    embedding_dimension: 384
+  model_id: all-MiniLM-L6-v2
+  provider_id: sentence-transformers
+  model_type: embedding
+shields:
+- shield_id: meta-llama/Llama-Guard-3-8B
+vector_dbs: []
+datasets: []
+scoring_fns: []
+benchmarks: []
+tool_groups:
+- toolgroup_id: builtin::websearch
+  provider_id: tavily-search
+- toolgroup_id: builtin::rag
+  provider_id: rag-runtime
+- toolgroup_id: builtin::code_interpreter
+  provider_id: code-interpreter
+server:
+  port: 8321

llama_stack/templates/verification/verification.py

@@ -0,0 +1,206 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from typing import List, Tuple
+
+from llama_stack.apis.models.models import ModelType
+from llama_stack.distribution.datatypes import (
+    ModelInput,
+    Provider,
+    ShieldInput,
+    ToolGroupInput,
+)
+from llama_stack.providers.inline.inference.sentence_transformers import (
+    SentenceTransformersInferenceConfig,
+)
+from llama_stack.providers.inline.vector_io.sqlite_vec.config import (
+    SQLiteVectorIOConfig,
+)
+from llama_stack.providers.remote.inference.cerebras.models import MODEL_ENTRIES as CEREBRAS_MODEL_ENTRIES
+from llama_stack.providers.remote.inference.cerebras_openai_compat.config import CerebrasCompatConfig
+from llama_stack.providers.remote.inference.fireworks.models import (
+    MODEL_ENTRIES as FIREWORKS_MODEL_ENTRIES,
+)
+from llama_stack.providers.remote.inference.fireworks_openai_compat.config import FireworksCompatConfig
+from llama_stack.providers.remote.inference.groq.models import (
+    MODEL_ENTRIES as GROQ_MODEL_ENTRIES,
+)
+from llama_stack.providers.remote.inference.groq_openai_compat.config import GroqCompatConfig
+from llama_stack.providers.remote.inference.openai.config import OpenAIConfig
+from llama_stack.providers.remote.inference.openai.models import (
+    MODEL_ENTRIES as OPENAI_MODEL_ENTRIES,
+)
+from llama_stack.providers.remote.inference.sambanova.models import MODEL_ENTRIES as SAMBANOVA_MODEL_ENTRIES
+from llama_stack.providers.remote.inference.sambanova_openai_compat.config import SambaNovaCompatConfig
+from llama_stack.providers.remote.inference.together.models import (
+    MODEL_ENTRIES as TOGETHER_MODEL_ENTRIES,
+)
+from llama_stack.providers.remote.inference.together_openai_compat.config import TogetherCompatConfig
+from llama_stack.providers.remote.vector_io.chroma.config import ChromaVectorIOConfig
+from llama_stack.providers.remote.vector_io.pgvector.config import (
+    PGVectorVectorIOConfig,
+)
+from llama_stack.templates.template import (
+    DistributionTemplate,
+    RunConfigSettings,
+    get_model_registry,
+)
+
+
+def get_inference_providers() -> Tuple[List[Provider], List[ModelInput]]:
+    # in this template, we allow each API key to be optional
+    providers = [
+        (
+            "openai",
+            OPENAI_MODEL_ENTRIES,
+            OpenAIConfig.sample_run_config(api_key="${env.OPENAI_API_KEY:}"),
+        ),
+        (
+            "fireworks-openai-compat",
+            FIREWORKS_MODEL_ENTRIES,
+            FireworksCompatConfig.sample_run_config(api_key="${env.FIREWORKS_API_KEY:}"),
+        ),
+        (
+            "together-openai-compat",
+            TOGETHER_MODEL_ENTRIES,
+            TogetherCompatConfig.sample_run_config(api_key="${env.TOGETHER_API_KEY:}"),
+        ),
+        (
+            "groq-openai-compat",
+            GROQ_MODEL_ENTRIES,
+            GroqCompatConfig.sample_run_config(api_key="${env.GROQ_API_KEY:}"),
+        ),
+        (
+            "sambanova-openai-compat",
+            SAMBANOVA_MODEL_ENTRIES,
+            SambaNovaCompatConfig.sample_run_config(api_key="${env.SAMBANOVA_API_KEY:}"),
+        ),
+        (
+            "cerebras-openai-compat",
+            CEREBRAS_MODEL_ENTRIES,
+            CerebrasCompatConfig.sample_run_config(api_key="${env.CEREBRAS_API_KEY:}"),
+        ),
+    ]
+    inference_providers = []
+    available_models = {}
+    for provider_id, model_entries, config in providers:
+        inference_providers.append(
+            Provider(
+                provider_id=provider_id,
+                provider_type=f"remote::{provider_id}",
+                config=config,
+            )
+        )
+        available_models[provider_id] = model_entries
+    return inference_providers, available_models
+
+
+def get_distribution_template() -> DistributionTemplate:
+    inference_providers, available_models = get_inference_providers()
+    providers = {
+        "inference": ([p.provider_type for p in inference_providers] + ["inline::sentence-transformers"]),
+        "vector_io": ["inline::sqlite-vec", "remote::chromadb", "remote::pgvector"],
+        "safety": ["inline::llama-guard"],
+        "agents": ["inline::meta-reference"],
+        "telemetry": ["inline::meta-reference"],
+        "eval": ["inline::meta-reference"],
+        "datasetio": ["remote::huggingface", "inline::localfs"],
+        "scoring": ["inline::basic", "inline::llm-as-judge", "inline::braintrust"],
+        "tool_runtime": [
+            "remote::brave-search",
+            "remote::tavily-search",
+            "inline::code-interpreter",
+            "inline::rag-runtime",
+            "remote::model-context-protocol",
+        ],
+    }
+    name = "verification"
+
+    vector_io_providers = [
+        Provider(
+            provider_id="sqlite-vec",
+            provider_type="inline::sqlite-vec",
+            config=SQLiteVectorIOConfig.sample_run_config(f"~/.llama/distributions/{name}"),
+        ),
+        Provider(
+            provider_id="${env.ENABLE_CHROMADB+chromadb}",
+            provider_type="remote::chromadb",
+            config=ChromaVectorIOConfig.sample_run_config(url="${env.CHROMADB_URL:}"),
+        ),
+        Provider(
+            provider_id="${env.ENABLE_PGVECTOR+pgvector}",
+            provider_type="remote::pgvector",
+            config=PGVectorVectorIOConfig.sample_run_config(
+                db="${env.PGVECTOR_DB:}",
+                user="${env.PGVECTOR_USER:}",
+                password="${env.PGVECTOR_PASSWORD:}",
+            ),
+        ),
+    ]
+    embedding_provider = Provider(
+        provider_id="sentence-transformers",
+        provider_type="inline::sentence-transformers",
+        config=SentenceTransformersInferenceConfig.sample_run_config(),
+    )
+
+    default_tool_groups = [
+        ToolGroupInput(
+            toolgroup_id="builtin::websearch",
+            provider_id="tavily-search",
+        ),
+        ToolGroupInput(
+            toolgroup_id="builtin::rag",
+            provider_id="rag-runtime",
+        ),
+        ToolGroupInput(
+            toolgroup_id="builtin::code_interpreter",
+            provider_id="code-interpreter",
+        ),
+    ]
+    embedding_model = ModelInput(
+        model_id="all-MiniLM-L6-v2",
+        provider_id=embedding_provider.provider_id,
+        model_type=ModelType.embedding,
+        metadata={
+            "embedding_dimension": 384,
+        },
+    )
+
+    default_models = get_model_registry(available_models)
+    return DistributionTemplate(
+        name=name,
+        distro_type="self_hosted",
+        description="Distribution for running e2e tests in CI",
+        container_image=None,
+        template_path=None,
+        providers=providers,
+        available_models_by_provider=available_models,
+        run_configs={
+            "run.yaml": RunConfigSettings(
+                provider_overrides={
+                    "inference": inference_providers + [embedding_provider],
+                    "vector_io": vector_io_providers,
+                },
+                default_models=default_models + [embedding_model],
+                default_tool_groups=default_tool_groups,
+                default_shields=[ShieldInput(shield_id="meta-llama/Llama-Guard-3-8B")],
+            ),
+        },
+        run_config_env_vars={
+            "LLAMA_STACK_PORT": (
+                "8321",
+                "Port for the Llama Stack distribution server",
+            ),
+            "FIREWORKS_API_KEY": (
+                "",
+                "Fireworks API Key",
+            ),
+            "OPENAI_API_KEY": (
+                "",
+                "OpenAI API Key",
+            ),
+        },
+    )

tests/integration/inference/test_text_inference.py

@@ -6,6 +6,7 @@
 
 
 import os
+from time import sleep
 
 import pytest
 from pydantic import BaseModel
@@ -23,11 +24,15 @@ def skip_if_model_doesnt_support_completion(client_with_models, model_id):
     provider_id = models[model_id].provider_id
     providers = {p.provider_id: p for p in client_with_models.providers.list()}
     provider = providers[provider_id]
-    if provider.provider_type in (
-        "remote::openai",
-        "remote::anthropic",
-        "remote::gemini",
-        "remote::groq",
+    if (
+        provider.provider_type
+        in (
+            "remote::openai",
+            "remote::anthropic",
+            "remote::gemini",
+            "remote::groq",
+        )
+        or "openai-compat" in provider.provider_type
     ):
         pytest.skip(f"Model {model_id} hosted by {provider.provider_type} doesn't support completion")
 
@@ -514,7 +519,7 @@ def test_text_chat_completion_tool_calling_tools_not_in_request(
 )
 def test_text_chat_completion_with_multi_turn_tool_calling(client_with_models, text_model_id, test_case):
     """This test tests the model's tool calling loop in various scenarios"""
-    if "llama-4" not in text_model_id.lower():
+    if "llama-4" not in text_model_id.lower() and "llama4" not in text_model_id.lower():
         pytest.xfail("Not tested for non-llama4 models yet")
 
     tc = TestCase(test_case)
@@ -545,7 +550,7 @@ def test_text_chat_completion_with_multi_turn_tool_calling(client_with_models, t
         )
         op_msg = response.completion_message
         messages.append(op_msg.model_dump())
-        # pprint(op_msg)
+        # print(op_msg)
 
         assert op_msg.role == "assistant"
         expected = tc["expected"].pop(0)
@@ -568,3 +573,6 @@ def test_text_chat_completion_with_multi_turn_tool_calling(client_with_models, t
             actual_answer = op_msg.content.lower()
             # pprint(actual_answer)
             assert expected["answer"] in actual_answer
+
+        # sleep to avoid rate limit
+        sleep(1)

tests/integration/inference/test_vision_inference.py

@@ -76,8 +76,9 @@ def multi_image_data():
 
 @pytest.mark.parametrize("stream", [True, False])
 def test_image_chat_completion_multiple_images(client_with_models, vision_model_id, multi_image_data, stream):
-    if "llama-4" not in vision_model_id.lower() and "gpt-4o" not in vision_model_id.lower():
-        pytest.skip("Skip for non-llama4, gpt4o models")
+    supported_models = ["llama-4", "gpt-4o", "llama4"]
+    if not any(model in vision_model_id.lower() for model in supported_models):
+        pytest.skip(f"Skip for non-supported model: {vision_model_id}")
 
     messages = [
         {

tests/integration/test_cases/inference/chat_completion.json

@@ -220,7 +220,7 @@
               "description": "Availability status of the product."
             },
             "tags": {
-              "param_type": "list",
+              "param_type": "list[str]",
               "description": "List of product tags"
             }
           }
@@ -294,7 +294,7 @@
               "description": "Location of the event"
             },
             "participants": {
-              "param_type": "list",
+              "param_type": "list[str]",
               "description": "List of participant names"
             }
           }