diff --git a/backends/arm/test/models/test_llama.py b/backends/arm/test/models/test_llama.py
index 2b524fe0cc9..84eec491c1e 100644
--- a/backends/arm/test/models/test_llama.py
+++ b/backends/arm/test/models/test_llama.py
@@ -22,13 +22,13 @@
     TosaPipelineBI,
     TosaPipelineMI,
 )
-
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.examples.models.llama.export_llama_lib import (
     build_args_parser,
     get_llama_model,
 )
 
+from executorch.extension.llm.export.config.llm_config import LlmConfig
+
 input_t = Tuple[torch.Tensor]
 
 # Add project dir to sys path to workaround importlib.import_module() conditions in model_factory.py
diff --git a/examples/apple/mps/scripts/mps_example.py b/examples/apple/mps/scripts/mps_example.py
index 4aa34fce9f8..42ea79435ed 100644
--- a/examples/apple/mps/scripts/mps_example.py
+++ b/examples/apple/mps/scripts/mps_example.py
@@ -19,8 +19,6 @@
 from executorch.devtools.bundled_program.serialize import (
     serialize_from_bundled_program_to_flatbuffer,
 )
-
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.exir import (
     EdgeCompileConfig,
     EdgeProgramManager,
@@ -31,6 +29,8 @@
 from executorch.exir.capture._config import ExecutorchBackendConfig
 from executorch.extension.export_util.utils import export_to_edge, save_pte_program
 
+from executorch.extension.llm.export.config.llm_config import LlmConfig
+
 from ....models import MODEL_NAME_TO_MODEL
 from ....models.model_factory import EagerModelFactory
 
diff --git a/examples/models/llama/TARGETS b/examples/models/llama/TARGETS
index d2caccd5897..95d57e12f5a 100644
--- a/examples/models/llama/TARGETS
+++ b/examples/models/llama/TARGETS
@@ -67,7 +67,7 @@ runtime.python_library(
         "//caffe2:torch",
         "//executorch/examples/models:model_base",
         "//executorch/examples/models/llama:llama_transformer",
-        "//executorch/examples/models/llama/config:llm_config",
+        "//executorch/extension/llm/export/config:llm_config",
         "//executorch/examples/models:checkpoint",
     ],
 )
@@ -150,7 +150,7 @@ runtime.python_library(
         ":source_transformation",
         "//ai_codesign/gen_ai/fast_hadamard_transform:fast_hadamard_transform",
         "//caffe2:torch",
-        "//executorch/examples/models/llama/config:llm_config",
+        "//executorch/extension/llm/export/config:llm_config",
         "//executorch/backends/vulkan/_passes:vulkan_passes",
         "//executorch/exir/passes:init_mutable_pass",
         "//executorch/examples/models:model_base",
diff --git a/examples/models/llama/config/targets.bzl b/examples/models/llama/config/targets.bzl
index 8b85ce6d107..a72ffa6c8e0 100644
--- a/examples/models/llama/config/targets.bzl
+++ b/examples/models/llama/config/targets.bzl
@@ -2,25 +2,12 @@ load("@fbsource//xplat/executorch/build:runtime_wrapper.bzl", "runtime")
 load("@fbcode_macros//build_defs:python_unittest.bzl", "python_unittest")
 
 def define_common_targets():
-    runtime.python_library(
-        name = "llm_config",
-        srcs = [
-            "llm_config.py",
-        ],
-        _is_external_target = True,
-        base_module = "executorch.examples.models.llama.config",
-        visibility = [
-            "//executorch/...",
-            "@EXECUTORCH_CLIENTS",
-        ],
-    )
-
     python_unittest(
         name = "test_llm_config",
         srcs = [
             "test_llm_config.py",
         ],
         deps = [
-            ":llm_config",
+            "//executorch/extension/llm/export/config:llm_config",
         ],
     )
diff --git a/examples/models/llama/config/test_llm_config.py b/examples/models/llama/config/test_llm_config.py
index 52b56d71a03..ec85e4c2e92 100644
--- a/examples/models/llama/config/test_llm_config.py
+++ b/examples/models/llama/config/test_llm_config.py
@@ -8,7 +8,7 @@
 
 import unittest
 
-from executorch.examples.models.llama.config.llm_config import (
+from executorch.extension.llm.export.config.llm_config import (
     BackendConfig,
     BaseConfig,
     CoreMLComputeUnit,
diff --git a/examples/models/llama/eval_llama_lib.py b/examples/models/llama/eval_llama_lib.py
index 20ba6dbaa9f..991ff72ae43 100644
--- a/examples/models/llama/eval_llama_lib.py
+++ b/examples/models/llama/eval_llama_lib.py
@@ -175,7 +175,7 @@ def gen_eval_wrapper(
     """
     # If llm_config is not provided, convert args to llm_config
     if llm_config is None:
-        from executorch.examples.models.llama.config.llm_config import LlmConfig
+        from executorch.extension.llm.export.config.llm_config import LlmConfig
 
         llm_config = LlmConfig.from_args(args)
 
@@ -306,7 +306,7 @@ def eval_llama(
     args: argparse.ArgumentParser,
 ) -> None:
     # Convert args to LlmConfig
-    from executorch.examples.models.llama.config.llm_config import LlmConfig
+    from executorch.extension.llm.export.config.llm_config import LlmConfig
 
     llm_config = LlmConfig.from_args(args)
 
@@ -340,7 +340,7 @@ def eval_llama_with_attention_sink(model_name: str, args: argparse.ArgumentParse
     This is mostly copied from https://github.com/mit-han-lab/streaming-llm/blob/main/examples/eval_long_ppl.py
     """
     # Convert args to LlmConfig
-    from executorch.examples.models.llama.config.llm_config import LlmConfig
+    from executorch.extension.llm.export.config.llm_config import LlmConfig
 
     llm_config = LlmConfig.from_args(args)
 
diff --git a/examples/models/llama/export_llama_lib.py b/examples/models/llama/export_llama_lib.py
index 685e9de9a2e..43ae595f797 100644
--- a/examples/models/llama/export_llama_lib.py
+++ b/examples/models/llama/export_llama_lib.py
@@ -27,8 +27,6 @@
 from executorch.devtools.backend_debug import print_delegation_info
 
 from executorch.devtools.etrecord import generate_etrecord as generate_etrecord_func
-
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.examples.models.llama.hf_download import (
     download_and_convert_hf_checkpoint,
 )
@@ -36,6 +34,8 @@
 
 from executorch.extension.llm.export.builder import DType, LLMEdgeManager
 
+from executorch.extension.llm.export.config.llm_config import LlmConfig
+
 from executorch.extension.llm.export.partitioner_lib import (
     get_coreml_partitioner,
     get_mps_partitioner,
diff --git a/examples/models/llama/model.py b/examples/models/llama/model.py
index efea80dde2f..27d41ac90cd 100644
--- a/examples/models/llama/model.py
+++ b/examples/models/llama/model.py
@@ -15,11 +15,11 @@
     get_checkpoint_dtype,
     get_default_model_resource_dir,
 )
-
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.examples.models.llama.llama_transformer import construct_transformer
 from executorch.examples.models.llama.model_args import ModelArgs
 from executorch.examples.models.llama.rope import Rope
+
+from executorch.extension.llm.export.config.llm_config import LlmConfig
 from torchao.utils import TorchAOBaseTensor
 
 try:
diff --git a/examples/models/llama/runner/eager.py b/examples/models/llama/runner/eager.py
index c55ad0eea28..7e662317509 100644
--- a/examples/models/llama/runner/eager.py
+++ b/examples/models/llama/runner/eager.py
@@ -9,8 +9,6 @@
 from typing import Optional, Type
 
 import torch
-
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.examples.models.llama.export_llama_lib import (
     _prepare_for_llama_export,
     build_args_parser as _build_args_parser,
@@ -18,6 +16,8 @@
 from executorch.examples.models.llama.runner.generation import LlamaRunner
 from executorch.extension.llm.export.builder import LLMEdgeManager
 
+from executorch.extension.llm.export.config.llm_config import LlmConfig
+
 
 class EagerLlamaRunner(LlamaRunner):
     """
diff --git a/examples/models/llama/tests/test_export_llama_lib.py b/examples/models/llama/tests/test_export_llama_lib.py
index f2ac9497604..172517207de 100644
--- a/examples/models/llama/tests/test_export_llama_lib.py
+++ b/examples/models/llama/tests/test_export_llama_lib.py
@@ -7,11 +7,11 @@
 import unittest
 
 from executorch.devtools.backend_debug import get_delegation_info
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.examples.models.llama.export_llama_lib import (
     _export_llama,
     build_args_parser,
 )
+from executorch.extension.llm.export.config.llm_config import LlmConfig
 
 UNWANTED_OPS = [
     "aten_permute_copy_default",
diff --git a/examples/models/llama3_2_vision/runner/eager.py b/examples/models/llama3_2_vision/runner/eager.py
index 5e68a43bf8e..1431901c050 100644
--- a/examples/models/llama3_2_vision/runner/eager.py
+++ b/examples/models/llama3_2_vision/runner/eager.py
@@ -8,7 +8,6 @@
 from typing import Optional
 
 import torch
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 
 from executorch.examples.models.llama.export_llama_lib import _prepare_for_llama_export
 from executorch.examples.models.llama.runner.eager import execute_runner
@@ -16,6 +15,7 @@
     TorchTuneLlamaRunner,
 )
 from executorch.extension.llm.export import LLMEdgeManager
+from executorch.extension.llm.export.config.llm_config import LlmConfig
 
 
 class EagerLlamaRunner(TorchTuneLlamaRunner):
diff --git a/examples/models/llava/export_llava.py b/examples/models/llava/export_llava.py
index 32b3ff448ac..e0580aa859a 100644
--- a/examples/models/llava/export_llava.py
+++ b/examples/models/llava/export_llava.py
@@ -16,7 +16,6 @@
     get_symmetric_quantization_config,
     XNNPACKQuantizer,
 )
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.examples.models.llama.export_llama_lib import (
     get_quantizer_and_quant_params,
 )
@@ -44,6 +43,7 @@
     HintBasedSymShapeEvalPass,
 )
 from executorch.extension.llm.export.builder import DType, LLMEdgeManager
+from executorch.extension.llm.export.config.llm_config import LlmConfig
 from executorch.util.activation_memory_profiler import generate_memory_trace
 from pytorch_tokenizers.llama2c import Llama2cTokenizer as Tokenizer
 from torch.export import Dim
diff --git a/extension/llm/export/README.md b/extension/llm/export/README.md
index e97b9e10462..f5d3b95bbf3 100644
--- a/extension/llm/export/README.md
+++ b/extension/llm/export/README.md
@@ -96,7 +96,7 @@ Please refer to the docs for some of our example suported models ([Llama](https:
 
 ## Configuration Options
 
-For a complete reference of all available configuration options, see the [LlmConfig class definition](../../../examples/models/llama/config/llm_config.py) which documents all supported parameters for base, model, export, quantization, backend, and debug configurations.
+For a complete reference of all available configuration options, see the [LlmConfig class definition](config/llm_config.py) which documents all supported parameters for base, model, export, quantization, backend, and debug configurations.
 
 ## Further Reading
 
diff --git a/extension/llm/export/config/TARGETS b/extension/llm/export/config/TARGETS
new file mode 100644
index 00000000000..bf8d13dcf37
--- /dev/null
+++ b/extension/llm/export/config/TARGETS
@@ -0,0 +1,5 @@
+load(":targets.bzl", "define_common_targets")
+
+oncall("executorch")
+
+define_common_targets()
\ No newline at end of file
diff --git a/examples/models/llama/config/llm_config.py b/extension/llm/export/config/llm_config.py
similarity index 100%
rename from examples/models/llama/config/llm_config.py
rename to extension/llm/export/config/llm_config.py
diff --git a/extension/llm/export/config/targets.bzl b/extension/llm/export/config/targets.bzl
new file mode 100644
index 00000000000..4135b336fbd
--- /dev/null
+++ b/extension/llm/export/config/targets.bzl
@@ -0,0 +1,15 @@
+load("@fbsource//xplat/executorch/build:runtime_wrapper.bzl", "runtime")
+
+def define_common_targets():
+    runtime.python_library(
+        name = "llm_config",
+        srcs = [
+            "llm_config.py",
+        ],
+        _is_external_target = True,
+        base_module = "executorch.extension.llm.export.config",
+        visibility = [
+            "//executorch/...",
+            "@EXECUTORCH_CLIENTS",
+        ],
+    )
\ No newline at end of file
diff --git a/extension/llm/export/export_llm.py b/extension/llm/export/export_llm.py
index e0467250a28..9841ae854a8 100644
--- a/extension/llm/export/export_llm.py
+++ b/extension/llm/export/export_llm.py
@@ -35,9 +35,9 @@
 from typing import Any, List, Tuple
 
 import hydra
-
-from executorch.examples.models.llama.config.llm_config import LlmConfig
 from executorch.examples.models.llama.export_llama_lib import export_llama
+
+from executorch.extension.llm.export.config.llm_config import LlmConfig
 from hydra.core.config_store import ConfigStore
 from omegaconf import OmegaConf