mypy-boto3-bedrock Type Stubs

1.42.83 · active · verified Sat Apr 11

mypy-boto3-bedrock provides comprehensive type annotations for the `boto3` Bedrock service. It is part of the `boto3-stubs` ecosystem, automatically generated by `mypy-boto3-builder`, ensuring compatibility with `boto3`'s official releases. This library enhances static analysis with tools like `mypy` and `pyright`, offering improved code completion and helping identify potential bugs in `boto3` Bedrock client usage. The version number of `mypy-boto3-bedrock` is synchronized with the corresponding `boto3` version it provides stubs for.

Warnings

breaking Starting with `mypy-boto3-builder` version 8.12.0 (which generated `mypy-boto3-bedrock` 1.42.83 and newer), support for Python 3.8 has been removed. Projects using these stubs must target Python 3.9 or newer.
Fix: Upgrade your project's Python interpreter to 3.9 or higher.
breaking All `mypy-boto3-*` packages, including `mypy-boto3-bedrock`, have migrated to PEP 561 compliant distribution. This changes how type checkers might discover and consume the stubs.
Fix: Ensure your `mypy` or `pyright` configuration is up-to-date and correctly set up to use PEP 561 compliant stub packages. Typically, no manual action is needed beyond installing the stubs correctly.
breaking Service-specific `TypeDef` naming conventions changed for clarity in `mypy-boto3-builder` 8.9.0. For example, `CreateDistributionRequestRequestTypeDef` became `CreateDistributionRequestTypeDef`. This affects direct imports of generated `TypeDef` names.
Fix: Review your explicit `TypeDef` imports and update them according to the new naming conventions. Consult the specific package's documentation for exact changes if a `TypeDef` name causes issues.
gotcha PyCharm users may experience performance issues with `Literal` overloads when using full `boto3-stubs` packages. The `boto3-stubs-lite` variants are recommended as a workaround for PyCharm until this issue is resolved.
Fix: For PyCharm, consider installing `boto3-stubs-lite[bedrock]` instead of `boto3-stubs[bedrock]` or `mypy-boto3-bedrock` directly, and explicitly annotate types. Example: `pip install 'boto3-stubs-lite[bedrock]'`.
gotcha Pylint might report 'undefined variable' errors when using type stubs without `TYPE_CHECKING` guards, as it attempts to evaluate imports at runtime.
Fix: Wrap your stub imports within a `if TYPE_CHECKING:` block to ensure they are only visible to type checkers and ignored by Pylint at runtime. The quickstart example demonstrates this pattern.

Install

pip install mypy-boto3-bedrock boto3 Install standalone stubs and boto3
pip install 'boto3-stubs[bedrock]' boto3 Install via boto3-stubs extras (recommended for multiple services)

Imports

BedrockClient

from mypy_boto3_bedrock.client import BedrockClient

AgreementStatusType
```
from mypy_boto3_bedrock.literals import AgreementStatusType
```
Example of importing a generated Literal type.
ModelEnforcementOutputTypeDef
```
from mypy_boto3_bedrock.type_defs import ModelEnforcementOutputTypeDef
```
Example of importing a generated TypedDict.

Quickstart

This quickstart demonstrates how to set up type checking for the `boto3` Bedrock client using `mypy-boto3-bedrock`. It explicitly type hints the `boto3` client with `BedrockClient` and shows how to use a generated `TypeDef` for a response object, enabling static analysis and IDE autocompletion for Bedrock API calls.

import os
from typing import TYPE_CHECKING

import boto3
from boto3.session import Session

# Conditionally import type stubs only during type checking
# This avoids a runtime dependency on mypy-boto3-bedrock
if TYPE_CHECKING:
    from mypy_boto3_bedrock.client import BedrockClient
    from mypy_boto3_bedrock.type_defs import ListFoundationModelsResponseTypeDef

def get_bedrock_client() -> 'BedrockClient':
    # Ensure AWS credentials are configured (e.g., via environment variables or AWS CLI config)
    # For this example, we use a default session.
    # In a real application, consider explicit credential management.
    session = Session(region_name=os.environ.get('AWS_REGION', 'us-east-1'))
    client: BedrockClient = session.client('bedrock')
    return client

def list_bedrock_foundation_models() -> None:
    client = get_bedrock_client()
    print(f"Client type: {type(client)}")
    try:
        response: ListFoundationModelsResponseTypeDef = client.list_foundation_models()
        print("Successfully listed Bedrock foundation models:")
        for model in response['modelSummaries']:
            print(f"  - {model['modelId']} ({model['modelName']})")
    except Exception as e:
        print(f"Error listing models: {e}")

if __name__ == '__main__':
    list_bedrock_foundation_models()

view raw JSON →