TokenUsage dataclass

The tokens used in a message API inference call.

Source code in src/aws_sdk_bedrock_runtime/models.py

@dataclass(kw_only=True)
class TokenUsage:
    """The tokens used in a message API inference call."""

    input_tokens: int
    """The number of tokens sent in the request to the model."""

    output_tokens: int
    """The number of tokens that the model generated for the request."""

    total_tokens: int
    """The total of input tokens and tokens generated by the model."""

    cache_read_input_tokens: int | None = None
    """The number of input tokens read from the cache for the request."""

    cache_write_input_tokens: int | None = None
    """The number of input tokens written to the cache for the request."""

    def serialize(self, serializer: ShapeSerializer):
        serializer.write_struct(_SCHEMA_TOKEN_USAGE, self)

    def serialize_members(self, serializer: ShapeSerializer):
        serializer.write_integer(
            _SCHEMA_TOKEN_USAGE.members["inputTokens"], self.input_tokens
        )
        serializer.write_integer(
            _SCHEMA_TOKEN_USAGE.members["outputTokens"], self.output_tokens
        )
        serializer.write_integer(
            _SCHEMA_TOKEN_USAGE.members["totalTokens"], self.total_tokens
        )
        if self.cache_read_input_tokens is not None:
            serializer.write_integer(
                _SCHEMA_TOKEN_USAGE.members["cacheReadInputTokens"],
                self.cache_read_input_tokens,
            )

        if self.cache_write_input_tokens is not None:
            serializer.write_integer(
                _SCHEMA_TOKEN_USAGE.members["cacheWriteInputTokens"],
                self.cache_write_input_tokens,
            )

    @classmethod
    def deserialize(cls, deserializer: ShapeDeserializer) -> Self:
        return cls(**cls.deserialize_kwargs(deserializer))

    @classmethod
    def deserialize_kwargs(cls, deserializer: ShapeDeserializer) -> dict[str, Any]:
        kwargs: dict[str, Any] = {}

        def _consumer(schema: Schema, de: ShapeDeserializer) -> None:
            match schema.expect_member_index():
                case 0:
                    kwargs["input_tokens"] = de.read_integer(
                        _SCHEMA_TOKEN_USAGE.members["inputTokens"]
                    )

                case 1:
                    kwargs["output_tokens"] = de.read_integer(
                        _SCHEMA_TOKEN_USAGE.members["outputTokens"]
                    )

                case 2:
                    kwargs["total_tokens"] = de.read_integer(
                        _SCHEMA_TOKEN_USAGE.members["totalTokens"]
                    )

                case 3:
                    kwargs["cache_read_input_tokens"] = de.read_integer(
                        _SCHEMA_TOKEN_USAGE.members["cacheReadInputTokens"]
                    )

                case 4:
                    kwargs["cache_write_input_tokens"] = de.read_integer(
                        _SCHEMA_TOKEN_USAGE.members["cacheWriteInputTokens"]
                    )

                case _:
                    logger.debug("Unexpected member schema: %s", schema)

        deserializer.read_struct(_SCHEMA_TOKEN_USAGE, consumer=_consumer)
        return kwargs

Attributes

cache_read_input_tokens class-attribute instance-attribute

cache_read_input_tokens: int | None = None

The number of input tokens read from the cache for the request.

cache_write_input_tokens class-attribute instance-attribute

cache_write_input_tokens: int | None = None

The number of input tokens written to the cache for the request.

input_tokens instance-attribute

input_tokens: int

The number of tokens sent in the request to the model.

output_tokens instance-attribute

output_tokens: int

The number of tokens that the model generated for the request.

total_tokens instance-attribute

total_tokens: int

The total of input tokens and tokens generated by the model.