release: 0.29.0

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "0.28.1"
+  ".": "0.29.0"
 }

.stats.yml

@@ -1,4 +1,4 @@
 configured_endpoints: 49
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/mixedbread%2Fmixedbread-c950bdc025566e8000354aef54c6f1c6c4ecbd92f9b6aca231701b8bebdf917e.yml
-openapi_spec_hash: d76b267c95385a291cdb9401a9bdb95e
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/mixedbread%2Fmixedbread-a2720b37997685cf019808be37ab86f7071f55ac09c80916293489f6d3644160.yml
+openapi_spec_hash: 61c24c03f1df5276b03e7bcce5266c21
 config_hash: 22524d9aa927566851636768f6a861cd

CHANGELOG.md

@@ -1,5 +1,19 @@
 # Changelog
 
+## 0.29.0 (2025-09-01)
+
+Full Changelog: [v0.28.1...v0.29.0](https://github.com/mixedbread-ai/mixedbread-python/compare/v0.28.1...v0.29.0)
+
+### Features
+
+* **api:** api update ([9869487](https://github.com/mixedbread-ai/mixedbread-python/commit/986948712a20dd99f1e5e3c40e8180ab11a727dd))
+
+
+### Chores
+
+* **internal:** add Sequence related utils ([cc3f90d](https://github.com/mixedbread-ai/mixedbread-python/commit/cc3f90d4f28cc313c2a20acd628149d56b7be6bd))
+* **internal:** update pyright exclude list ([8d57f64](https://github.com/mixedbread-ai/mixedbread-python/commit/8d57f64f771bfec95fad5f0d481ce4538fabeb1e))
+
 ## 0.28.1 (2025-08-27)
 
 Full Changelog: [v0.28.0...v0.28.1](https://github.com/mixedbread-ai/mixedbread-python/compare/v0.28.0...v0.28.1)

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mixedbread"
-version = "0.28.1"
+version = "0.29.0"
 description = "The official Python library for the Mixedbread API"
 dynamic = ["readme"]
 license = "Apache-2.0"
@@ -148,6 +148,7 @@ exclude = [
     "_dev",
     ".venv",
     ".nox",
+    ".git",
 ]
 
 reportImplicitOverride = true

src/mixedbread/_types.py

@@ -13,10 +13,21 @@
     Mapping,
     TypeVar,
     Callable,
+    Iterator,
     Optional,
     Sequence,
 )
-from typing_extensions import Set, Literal, Protocol, TypeAlias, TypedDict, override, runtime_checkable
+from typing_extensions import (
+    Set,
+    Literal,
+    Protocol,
+    TypeAlias,
+    TypedDict,
+    SupportsIndex,
+    overload,
+    override,
+    runtime_checkable,
+)
 
 import httpx
 import pydantic
@@ -217,3 +228,26 @@ class _GenericAlias(Protocol):
 class HttpxSendArgs(TypedDict, total=False):
     auth: httpx.Auth
     follow_redirects: bool
+
+
+_T_co = TypeVar("_T_co", covariant=True)
+
+
+if TYPE_CHECKING:
+    # This works because str.__contains__ does not accept object (either in typeshed or at runtime)
+    # https://github.com/hauntsaninja/useful_types/blob/5e9710f3875107d068e7679fd7fec9cfab0eff3b/useful_types/__init__.py#L285
+    class SequenceNotStr(Protocol[_T_co]):
+        @overload
+        def __getitem__(self, index: SupportsIndex, /) -> _T_co: ...
+        @overload
+        def __getitem__(self, index: slice, /) -> Sequence[_T_co]: ...
+        def __contains__(self, value: object, /) -> bool: ...
+        def __len__(self) -> int: ...
+        def __iter__(self) -> Iterator[_T_co]: ...
+        def index(self, value: Any, start: int = 0, stop: int = ..., /) -> int: ...
+        def count(self, value: Any, /) -> int: ...
+        def __reversed__(self) -> Iterator[_T_co]: ...
+else:
+    # just point this to a normal `Sequence` at runtime to avoid having to special case
+    # deserializing our custom sequence type
+    SequenceNotStr = Sequence

src/mixedbread/_utils/__init__.py

@@ -38,6 +38,7 @@
     extract_type_arg as extract_type_arg,
     is_iterable_type as is_iterable_type,
     is_required_type as is_required_type,
+    is_sequence_type as is_sequence_type,
     is_annotated_type as is_annotated_type,
     is_type_alias_type as is_type_alias_type,
     strip_annotated_type as strip_annotated_type,

src/mixedbread/_utils/_typing.py

@@ -26,6 +26,11 @@ def is_list_type(typ: type) -> bool:
     return (get_origin(typ) or typ) == list
 
 
+def is_sequence_type(typ: type) -> bool:
+    origin = get_origin(typ) or typ
+    return origin == typing_extensions.Sequence or origin == typing.Sequence or origin == _c_abc.Sequence
+
+
 def is_iterable_type(typ: type) -> bool:
     """If the given type is `typing.Iterable[T]`"""
     origin = get_origin(typ) or typ

src/mixedbread/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "mixedbread"
-__version__ = "0.28.1"  # x-release-please-version
+__version__ = "0.29.0"  # x-release-please-version

src/mixedbread/types/vector_stores/vector_store_file.py

@@ -12,9 +12,12 @@
     "VectorStoreFile",
     "Chunk",
     "ChunkTextInputChunk",
-    "ChunkImageURLInputChunkBase",
-    "ChunkAudioURLInputChunkBase",
-    "ChunkVideoURLInputChunkBase",
+    "ChunkImageURLInputChunk",
+    "ChunkImageURLInputChunkImageURL",
+    "ChunkAudioURLInputChunk",
+    "ChunkAudioURLInputChunkAudioURL",
+    "ChunkVideoURLInputChunk",
+    "ChunkVideoURLInputChunkVideoURL",
 ]
 
 
@@ -41,7 +44,15 @@ class ChunkTextInputChunk(BaseModel):
     """Text content to process"""
 
 
-class ChunkImageURLInputChunkBase(BaseModel):
+class ChunkImageURLInputChunkImageURL(BaseModel):
+    url: str
+    """The image URL. Can be either a URL or a Data URI."""
+
+    format: Optional[str] = None
+    """The image format/mimetype"""
+
+
+class ChunkImageURLInputChunk(BaseModel):
     chunk_index: int
     """position of the chunk in a file"""
 
@@ -63,8 +74,16 @@ class ChunkImageURLInputChunkBase(BaseModel):
     summary: Optional[str] = None
     """summary of the image"""
 
+    image_url: ChunkImageURLInputChunkImageURL
+    """The image input specification."""
+
+
+class ChunkAudioURLInputChunkAudioURL(BaseModel):
+    url: str
+    """The audio URL. Can be either a URL or a Data URI."""
 
-class ChunkAudioURLInputChunkBase(BaseModel):
+
+class ChunkAudioURLInputChunk(BaseModel):
     chunk_index: int
     """position of the chunk in a file"""
 
@@ -86,8 +105,16 @@ class ChunkAudioURLInputChunkBase(BaseModel):
     summary: Optional[str] = None
     """summary of the audio"""
 
+    audio_url: ChunkAudioURLInputChunkAudioURL
+    """The audio input specification."""
+
+
+class ChunkVideoURLInputChunkVideoURL(BaseModel):
+    url: str
+    """The video URL. Can be either a URL or a Data URI."""
 
-class ChunkVideoURLInputChunkBase(BaseModel):
+
+class ChunkVideoURLInputChunk(BaseModel):
     chunk_index: int
     """position of the chunk in a file"""
 
@@ -109,9 +136,12 @@ class ChunkVideoURLInputChunkBase(BaseModel):
     summary: Optional[str] = None
     """summary of the video"""
 
+    video_url: ChunkVideoURLInputChunkVideoURL
+    """The video input specification."""
+
 
 Chunk: TypeAlias = Annotated[
-    Union[ChunkTextInputChunk, ChunkImageURLInputChunkBase, ChunkAudioURLInputChunkBase, ChunkVideoURLInputChunkBase],
+    Union[ChunkTextInputChunk, ChunkImageURLInputChunk, ChunkAudioURLInputChunk, ChunkVideoURLInputChunk],
     PropertyInfo(discriminator="type"),
 ]
 

tests/utils.py

@@ -4,7 +4,7 @@
 import inspect
 import traceback
 import contextlib
-from typing import Any, TypeVar, Iterator, cast
+from typing import Any, TypeVar, Iterator, Sequence, cast
 from datetime import date, datetime
 from typing_extensions import Literal, get_args, get_origin, assert_type
 
@@ -15,6 +15,7 @@
     is_list_type,
     is_union_type,
     extract_type_arg,
+    is_sequence_type,
     is_annotated_type,
     is_type_alias_type,
 )
@@ -71,6 +72,13 @@ def assert_matches_type(
     if is_list_type(type_):
         return _assert_list_type(type_, value)
 
+    if is_sequence_type(type_):
+        assert isinstance(value, Sequence)
+        inner_type = get_args(type_)[0]
+        for entry in value:  # type: ignore
+            assert_type(inner_type, entry)  # type: ignore
+        return
+
     if origin == str:
         assert isinstance(value, str)
     elif origin == int: