Skip to content

ImageBase64Content

Base classes

Name Children Inherits
BaseImageContent
llmling_agent.models.content
Base for image content.

⋔ Inheritance diagram

graph TD
  94662092174288["content.ImageBase64Content"]
  94662092188320["content.BaseImageContent"]
  94662092465904["content.BaseContent"]
  94662059216288["schema.Schema"]
  94662046694544["main.BaseModel"]
  140509591771616["builtins.object"]
  94662092188320 --> 94662092174288
  94662092465904 --> 94662092188320
  94662059216288 --> 94662092465904
  94662046694544 --> 94662059216288
  140509591771616 --> 94662046694544

🛈 DocStrings

Bases: BaseImageContent

Image from base64 data.

Source code in src/llmling_agent/models/content.py 
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
class ImageBase64Content(BaseImageContent):
    """Image from base64 data."""

    type: Literal["image_base64"] = Field("image_base64", init=False)
    """Base64-encoded image."""

    data: str
    """Base64-encoded image data."""

    mime_type: str = "image/jpeg"
    """MIME type of the image."""

    def to_pydantic_ai(self) -> BinaryImage:
        binary_data = base64.b64decode(self.data)
        return BinaryImage(data=binary_data, media_type=self.mime_type or "image/jpeg")

    @classmethod
    def from_bytes(
        cls,
        data: bytes,
        *,
        detail: DetailLevel | None = None,
        description: str | None = None,
    ) -> ImageBase64Content:
        """Create image content from raw bytes.

        Args:
            data: Raw image bytes
            detail: Optional detail level for processing
            description: Optional description of the image
        """
        content = base64.b64encode(data).decode()
        return cls(data=content, detail=detail, description=description)

    @classmethod
    def from_pil_image(cls, image: PIL.Image.Image) -> ImageBase64Content:
        """Create content from PIL Image."""
        with io.BytesIO() as buffer:
            image.save(buffer, format="PNG")
            return cls(data=base64.b64encode(buffer.getvalue()).decode())

data instance-attribute 

data: str

Base64-encoded image data.

mime_type class-attribute instance-attribute 

mime_type: str = 'image/jpeg'

MIME type of the image.

type class-attribute instance-attribute 

type: Literal['image_base64'] = Field('image_base64', init=False)

Base64-encoded image.

from_bytes classmethod 

from_bytes(
    data: bytes, *, detail: DetailLevel | None = None, description: str | None = None
) -> ImageBase64Content

Create image content from raw bytes.

Parameters:

Name Type Description Default
data bytes

Raw image bytes

 required
detail DetailLevel | None

Optional detail level for processing

 None
description str | None

Optional description of the image

 None
Source code in src/llmling_agent/models/content.py 
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
@classmethod
def from_bytes(
    cls,
    data: bytes,
    *,
    detail: DetailLevel | None = None,
    description: str | None = None,
) -> ImageBase64Content:
    """Create image content from raw bytes.

    Args:
        data: Raw image bytes
        detail: Optional detail level for processing
        description: Optional description of the image
    """
    content = base64.b64encode(data).decode()
    return cls(data=content, detail=detail, description=description)

from_pil_image classmethod 

from_pil_image(image: Image) -> ImageBase64Content

Create content from PIL Image.

Source code in src/llmling_agent/models/content.py 
137
138
139
140
141
142
@classmethod
def from_pil_image(cls, image: PIL.Image.Image) -> ImageBase64Content:
    """Create content from PIL Image."""
    with io.BytesIO() as buffer:
        image.save(buffer, format="PNG")
        return cls(data=base64.b64encode(buffer.getvalue()).decode())

Show source on GitHub