diff --git a/.release-please-manifest.json b/.release-please-manifest.json
index bf42d235..0dd91f48 100644
--- a/.release-please-manifest.json
+++ b/.release-please-manifest.json
@@ -1,3 +1,3 @@
 {
-  ".": "0.16.6"
+  ".": "0.16.7"
 }
\ No newline at end of file
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 8563af8c..debebc1f 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,15 @@
 # Changelog
 
+## 0.16.7 (2024-03-08)
+
+Full Changelog: [v0.16.6...v0.16.7](https://github.com/Finch-API/finch-api-python/compare/v0.16.6...v0.16.7)
+
+### Chores
+
+* export NOT_GIVEN sentinel value ([#321](https://github.com/Finch-API/finch-api-python/issues/321)) ([3091f44](https://github.com/Finch-API/finch-api-python/commit/3091f4411fe93b70ae8b2dcdeca9dae9f35b46ce))
+* **internal:** bump pyright ([#318](https://github.com/Finch-API/finch-api-python/issues/318)) ([0822881](https://github.com/Finch-API/finch-api-python/commit/082288138c7216f8ed7c4d499109201cf3aa626a))
+* **internal:** support parsing Annotated types ([#320](https://github.com/Finch-API/finch-api-python/issues/320)) ([6d2c896](https://github.com/Finch-API/finch-api-python/commit/6d2c896094ebcaf5a95a78d5d50bfe6094cc7305))
+
 ## 0.16.6 (2024-03-06)
 
 Full Changelog: [v0.16.5...v0.16.6](https://github.com/Finch-API/finch-api-python/compare/v0.16.5...v0.16.6)
diff --git a/pyproject.toml b/pyproject.toml
index 25539869..d8803cfc 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [project]
 name = "finch-api"
-version = "0.16.6"
+version = "0.16.7"
 description = "The official Python library for the Finch API"
 readme = "README.md"
 license = "Apache-2.0"
diff --git a/requirements-dev.lock b/requirements-dev.lock
index b00261be..672caedf 100644
--- a/requirements-dev.lock
+++ b/requirements-dev.lock
@@ -63,7 +63,7 @@ pydantic==2.4.2
     # via finch-api
 pydantic-core==2.10.1
     # via pydantic
-pyright==1.1.351
+pyright==1.1.353
 pytest==7.1.1
     # via pytest-asyncio
 pytest-asyncio==0.21.1
diff --git a/src/finch/__init__.py b/src/finch/__init__.py
index bdd86516..275a5e1f 100644
--- a/src/finch/__init__.py
+++ b/src/finch/__init__.py
@@ -1,7 +1,7 @@
 # File generated from our OpenAPI spec by Stainless.
 
 from . import types
-from ._types import NoneType, Transport, ProxiesTypes
+from ._types import NOT_GIVEN, NoneType, NotGiven, Transport, ProxiesTypes
 from ._utils import file_from_path
 from ._client import Finch, Client, Stream, Timeout, Transport, AsyncFinch, AsyncClient, AsyncStream, RequestOptions
 from ._models import BaseModel
@@ -32,6 +32,8 @@
     "NoneType",
     "Transport",
     "ProxiesTypes",
+    "NotGiven",
+    "NOT_GIVEN",
     "FinchError",
     "APIError",
     "APIStatusError",
diff --git a/src/finch/_legacy_response.py b/src/finch/_legacy_response.py
index f2a50c85..c3911ba8 100644
--- a/src/finch/_legacy_response.py
+++ b/src/finch/_legacy_response.py
@@ -13,7 +13,7 @@
 import pydantic
 
 from ._types import NoneType
-from ._utils import is_given
+from ._utils import is_given, extract_type_arg, is_annotated_type
 from ._models import BaseModel, is_basemodel
 from ._constants import RAW_RESPONSE_HEADER
 from ._streaming import Stream, AsyncStream, is_stream_class_type, extract_stream_chunk_type
@@ -174,6 +174,10 @@ def elapsed(self) -> datetime.timedelta:
         return self.http_response.elapsed
 
     def _parse(self, *, to: type[_T] | None = None) -> R | _T:
+        # unwrap `Annotated[T, ...]` -> `T`
+        if to and is_annotated_type(to):
+            to = extract_type_arg(to, 0)
+
         if self._stream:
             if to:
                 if not is_stream_class_type(to):
@@ -215,6 +219,11 @@ def _parse(self, *, to: type[_T] | None = None) -> R | _T:
             )
 
         cast_to = to if to is not None else self._cast_to
+
+        # unwrap `Annotated[T, ...]` -> `T`
+        if is_annotated_type(cast_to):
+            cast_to = extract_type_arg(cast_to, 0)
+
         if cast_to is NoneType:
             return cast(R, None)
 
@@ -315,7 +324,7 @@ def to_raw_response_wrapper(func: Callable[P, R]) -> Callable[P, LegacyAPIRespon
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> LegacyAPIResponse[R]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, str] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "true"
 
         kwargs["extra_headers"] = extra_headers
@@ -332,7 +341,7 @@ def async_to_raw_response_wrapper(func: Callable[P, Awaitable[R]]) -> Callable[P
 
     @functools.wraps(func)
     async def wrapped(*args: P.args, **kwargs: P.kwargs) -> LegacyAPIResponse[R]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, str] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "true"
 
         kwargs["extra_headers"] = extra_headers
diff --git a/src/finch/_models.py b/src/finch/_models.py
index 81089149..af68e618 100644
--- a/src/finch/_models.py
+++ b/src/finch/_models.py
@@ -30,7 +30,16 @@
     AnyMapping,
     HttpxRequestFiles,
 )
-from ._utils import is_list, is_given, is_mapping, parse_date, parse_datetime, strip_not_given
+from ._utils import (
+    is_list,
+    is_given,
+    is_mapping,
+    parse_date,
+    parse_datetime,
+    strip_not_given,
+    extract_type_arg,
+    is_annotated_type,
+)
 from ._compat import (
     PYDANTIC_V2,
     ConfigDict,
@@ -275,6 +284,9 @@ def construct_type(*, value: object, type_: type) -> object:
 
     If the given value does not match the expected type then it is returned as-is.
     """
+    # unwrap `Annotated[T, ...]` -> `T`
+    if is_annotated_type(type_):
+        type_ = extract_type_arg(type_, 0)
 
     # we need to use the origin class for any types that are subscripted generics
     # e.g. Dict[str, object]
diff --git a/src/finch/_response.py b/src/finch/_response.py
index 240b3f6c..df9358c9 100644
--- a/src/finch/_response.py
+++ b/src/finch/_response.py
@@ -25,7 +25,7 @@
 import pydantic
 
 from ._types import NoneType
-from ._utils import is_given, extract_type_var_from_base
+from ._utils import is_given, extract_type_arg, is_annotated_type, extract_type_var_from_base
 from ._models import BaseModel, is_basemodel
 from ._constants import RAW_RESPONSE_HEADER, OVERRIDE_CAST_TO_HEADER
 from ._streaming import Stream, AsyncStream, is_stream_class_type, extract_stream_chunk_type
@@ -121,6 +121,10 @@ def __repr__(self) -> str:
         )
 
     def _parse(self, *, to: type[_T] | None = None) -> R | _T:
+        # unwrap `Annotated[T, ...]` -> `T`
+        if to and is_annotated_type(to):
+            to = extract_type_arg(to, 0)
+
         if self._is_sse_stream:
             if to:
                 if not is_stream_class_type(to):
@@ -162,6 +166,11 @@ def _parse(self, *, to: type[_T] | None = None) -> R | _T:
             )
 
         cast_to = to if to is not None else self._cast_to
+
+        # unwrap `Annotated[T, ...]` -> `T`
+        if is_annotated_type(cast_to):
+            cast_to = extract_type_arg(cast_to, 0)
+
         if cast_to is NoneType:
             return cast(R, None)
 
@@ -634,7 +643,7 @@ def to_streamed_response_wrapper(func: Callable[P, R]) -> Callable[P, ResponseCo
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> ResponseContextManager[APIResponse[R]]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, str] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "stream"
 
         kwargs["extra_headers"] = extra_headers
@@ -655,7 +664,7 @@ def async_to_streamed_response_wrapper(
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> AsyncResponseContextManager[AsyncAPIResponse[R]]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, str] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "stream"
 
         kwargs["extra_headers"] = extra_headers
@@ -679,7 +688,7 @@ def to_custom_streamed_response_wrapper(
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> ResponseContextManager[_APIResponseT]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, Any] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "stream"
         extra_headers[OVERRIDE_CAST_TO_HEADER] = response_cls
 
@@ -704,7 +713,7 @@ def async_to_custom_streamed_response_wrapper(
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> AsyncResponseContextManager[_AsyncAPIResponseT]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, Any] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "stream"
         extra_headers[OVERRIDE_CAST_TO_HEADER] = response_cls
 
@@ -724,7 +733,7 @@ def to_raw_response_wrapper(func: Callable[P, R]) -> Callable[P, APIResponse[R]]
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> APIResponse[R]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, str] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "raw"
 
         kwargs["extra_headers"] = extra_headers
@@ -741,7 +750,7 @@ def async_to_raw_response_wrapper(func: Callable[P, Awaitable[R]]) -> Callable[P
 
     @functools.wraps(func)
     async def wrapped(*args: P.args, **kwargs: P.kwargs) -> AsyncAPIResponse[R]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, str] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "raw"
 
         kwargs["extra_headers"] = extra_headers
@@ -763,7 +772,7 @@ def to_custom_raw_response_wrapper(
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> _APIResponseT:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, Any] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "raw"
         extra_headers[OVERRIDE_CAST_TO_HEADER] = response_cls
 
@@ -786,7 +795,7 @@ def async_to_custom_raw_response_wrapper(
 
     @functools.wraps(func)
     def wrapped(*args: P.args, **kwargs: P.kwargs) -> Awaitable[_AsyncAPIResponseT]:
-        extra_headers = {**(cast(Any, kwargs.get("extra_headers")) or {})}
+        extra_headers: dict[str, Any] = {**(cast(Any, kwargs.get("extra_headers")) or {})}
         extra_headers[RAW_RESPONSE_HEADER] = "raw"
         extra_headers[OVERRIDE_CAST_TO_HEADER] = response_cls
 
diff --git a/src/finch/_version.py b/src/finch/_version.py
index 620249f5..3911f730 100644
--- a/src/finch/_version.py
+++ b/src/finch/_version.py
@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless.
 
 __title__ = "finch"
-__version__ = "0.16.6"  # x-release-please-version
+__version__ = "0.16.7"  # x-release-please-version
diff --git a/tests/test_legacy_response.py b/tests/test_legacy_response.py
index d86dc9ed..994448d4 100644
--- a/tests/test_legacy_response.py
+++ b/tests/test_legacy_response.py
@@ -1,4 +1,6 @@
 import json
+from typing import cast
+from typing_extensions import Annotated
 
 import httpx
 import pytest
@@ -63,3 +65,20 @@ def test_response_parse_custom_model(client: Finch) -> None:
     obj = response.parse(to=CustomModel)
     assert obj.foo == "hello!"
     assert obj.bar == 2
+
+
+def test_response_parse_annotated_type(client: Finch) -> None:
+    response = LegacyAPIResponse(
+        raw=httpx.Response(200, content=json.dumps({"foo": "hello!", "bar": 2})),
+        client=client,
+        stream=False,
+        stream_cls=None,
+        cast_to=str,
+        options=FinalRequestOptions.construct(method="get", url="/foo"),
+    )
+
+    obj = response.parse(
+        to=cast("type[CustomModel]", Annotated[CustomModel, "random metadata"]),
+    )
+    assert obj.foo == "hello!"
+    assert obj.bar == 2
diff --git a/tests/test_models.py b/tests/test_models.py
index ecdf47a1..7e6131e7 100644
--- a/tests/test_models.py
+++ b/tests/test_models.py
@@ -1,14 +1,14 @@
 import json
 from typing import Any, Dict, List, Union, Optional, cast
 from datetime import datetime, timezone
-from typing_extensions import Literal
+from typing_extensions import Literal, Annotated
 
 import pytest
 import pydantic
 from pydantic import Field
 
 from finch._compat import PYDANTIC_V2, parse_obj, model_dump, model_json
-from finch._models import BaseModel
+from finch._models import BaseModel, construct_type
 
 
 class BasicModel(BaseModel):
@@ -571,3 +571,15 @@ class OurModel(BaseModel):
         foo: Optional[str] = None
 
     takes_pydantic(OurModel())
+
+
+def test_annotated_types() -> None:
+    class Model(BaseModel):
+        value: str
+
+    m = construct_type(
+        value={"value": "foo"},
+        type_=cast(Any, Annotated[Model, "random metadata"]),
+    )
+    assert isinstance(m, Model)
+    assert m.value == "foo"
diff --git a/tests/test_response.py b/tests/test_response.py
index ba4ccb28..3e70ad63 100644
--- a/tests/test_response.py
+++ b/tests/test_response.py
@@ -1,5 +1,6 @@
 import json
-from typing import List
+from typing import List, cast
+from typing_extensions import Annotated
 
 import httpx
 import pytest
@@ -157,3 +158,37 @@ async def test_async_response_parse_custom_model(async_client: AsyncFinch) -> No
     obj = await response.parse(to=CustomModel)
     assert obj.foo == "hello!"
     assert obj.bar == 2
+
+
+def test_response_parse_annotated_type(client: Finch) -> None:
+    response = APIResponse(
+        raw=httpx.Response(200, content=json.dumps({"foo": "hello!", "bar": 2})),
+        client=client,
+        stream=False,
+        stream_cls=None,
+        cast_to=str,
+        options=FinalRequestOptions.construct(method="get", url="/foo"),
+    )
+
+    obj = response.parse(
+        to=cast("type[CustomModel]", Annotated[CustomModel, "random metadata"]),
+    )
+    assert obj.foo == "hello!"
+    assert obj.bar == 2
+
+
+async def test_async_response_parse_annotated_type(async_client: AsyncFinch) -> None:
+    response = AsyncAPIResponse(
+        raw=httpx.Response(200, content=json.dumps({"foo": "hello!", "bar": 2})),
+        client=async_client,
+        stream=False,
+        stream_cls=None,
+        cast_to=str,
+        options=FinalRequestOptions.construct(method="get", url="/foo"),
+    )
+
+    obj = await response.parse(
+        to=cast("type[CustomModel]", Annotated[CustomModel, "random metadata"]),
+    )
+    assert obj.foo == "hello!"
+    assert obj.bar == 2
diff --git a/tests/utils.py b/tests/utils.py
index 3217f048..7e3d3203 100644
--- a/tests/utils.py
+++ b/tests/utils.py
@@ -14,6 +14,8 @@
     is_list,
     is_list_type,
     is_union_type,
+    extract_type_arg,
+    is_annotated_type,
 )
 from finch._compat import PYDANTIC_V2, field_outer_type, get_model_fields
 from finch._models import BaseModel
@@ -49,6 +51,10 @@ def assert_matches_type(
     path: list[str],
     allow_none: bool = False,
 ) -> None:
+    # unwrap `Annotated[T, ...]` -> `T`
+    if is_annotated_type(type_):
+        type_ = extract_type_arg(type_, 0)
+
     if allow_none and value is None:
         return