Type Annotations for boto3 Bedrock Runtime

1.42.82 · active · verified Fri Apr 17

This package provides comprehensive type annotations (stubs) for the `boto3` Bedrock Runtime client, enabling static type checking with tools like `mypy`. It ensures that your interactions with the AWS Bedrock Runtime service are type-safe, catching potential errors at development time. The current version is 1.42.82, and new releases are frequent, typically aligning with `boto3` and `botocore` updates.

Common errors

```
ModuleNotFoundError: No module named 'types_boto3_bedrock_runtime'
```
cause Attempting to import the stub package directly at runtime, or the package is not installed.

fix Ensure the package is installed (`pip install types-boto3-bedrock-runtime`) and wrap any imports of types with `from typing import TYPE_CHECKING` and `if TYPE_CHECKING:`.
```
TypeError: Argument of type 'dict[str, Any]' cannot be assigned to parameter of type 'InvokeModelRequestRequestTypeDef' in call
```
cause Your dictionary for a request payload does not exactly match the structure defined by the `TypeDef` in the type stubs, or you forgot to pass the required keyword arguments.

fix Inspect the `InvokeModelRequestRequestTypeDef` (or similar `TypeDef`) definition to ensure your dictionary keys and types match precisely. `mypy` will point out the specific mismatch.
```
error: Library "boto3" has no py.typed file (hint: see https://mypy.readthedocs.io/en/stable/running_mypy.html#missing-library-stubs)
```
cause You are using `boto3` without its core type stubs (`types-boto3`) or without the correct `mypy` configuration to find these stubs.

fix Install the base `types-boto3` package (`pip install types-boto3`) in addition to `types-boto3-bedrock-runtime`, and ensure `mypy` is configured to include stub paths if necessary (though usually automatic).

Warnings

breaking Starting with version 8.12.0, `types-boto3-bedrock-runtime` and other `mypy-boto3-builder` packages migrated to PEP 561 standard packages. This means older `mypy` configurations might need adjustment, though most modern setups will handle it seamlessly. Python 3.8 support was also removed.
Fix: Ensure your `mypy` configuration is up-to-date. If using Python 3.8, you must use an older version of `types-boto3-bedrock-runtime` or upgrade to Python 3.9+.
breaking In version 8.9.0, there were breaking changes to how TypeDef names are generated, leading to shorter names (e.g., `CreateDistributionRequestRequestTypeDef` -> `CreateDistributionRequestTypeDef`). While this example is not for Bedrock Runtime, it indicates a pattern that might affect specific Bedrock Runtime TypeDefs in future updates if their names become redundant.
Fix: If encountering `NameError` or `ImportError` for TypeDefs, check the changelog for the specific version for renaming conventions and adjust your imports/references accordingly.
gotcha This package provides only type stubs for static analysis and does not contain any runtime code. You must have the actual `boto3` library installed for your application to run.
Fix: Always install `boto3` alongside `types-boto3-bedrock-runtime`: `pip install boto3 types-boto3-bedrock-runtime`.
gotcha Directly importing from `types_boto3_bedrock_runtime` at runtime (e.g., `from types_boto3_bedrock_runtime.client import BedrockRuntimeClient`) can lead to `ModuleNotFoundError` or other issues if the stub package is not designed for runtime import, or simply adds unnecessary overhead. The stubs are meant for static checkers like `mypy`.
Fix: Always wrap imports of types from `types-boto3-*` packages within an `if TYPE_CHECKING:` block. This ensures they are only considered by type checkers and are skipped at runtime.

Install

pip install types-boto3-bedrock-runtime boto3 mypy Install package and dependencies

Imports

BedrockRuntimeClient

wrong:

from types_boto3_bedrock_runtime import BedrockRuntimeClient

correct:

from typing import TYPE_CHECKING
if TYPE_CHECKING:
    from types_boto3_bedrock_runtime.client import BedrockRuntimeClient

Type stub packages should only be imported for type checking purposes to avoid runtime overhead or errors. Use `if TYPE_CHECKING:`.

InvokeModelRequestRequestTypeDef
```
from typing import TYPE_CHECKING
if TYPE_CHECKING:
    from types_boto3_bedrock_runtime.type_defs import InvokeModelRequestRequestTypeDef
```
Specific type definitions for requests/responses are found in `type_defs` and should also be conditionally imported.

Quickstart

This quickstart demonstrates how to initialize a `boto3` Bedrock Runtime client with type annotations. It shows the conditional import of `BedrockRuntimeClient` and `InvokeModelRequestRequestTypeDef` for static analysis and then uses the client to invoke a model. Remember to set your AWS credentials and region via environment variables or other boto3 configuration methods for actual execution.

import boto3
import os
from typing import TYPE_CHECKING

# Conditional import for type checking
if TYPE_CHECKING:
    from types_boto3_bedrock_runtime.client import BedrockRuntimeClient
    from types_boto3_bedrock_runtime.type_defs import InvokeModelRequestRequestTypeDef


def invoke_bedrock_model(model_id: str, prompt: str):
    # Type annotate the client
    client: BedrockRuntimeClient = boto3.client(
        'bedrock-runtime',
        aws_access_key_id=os.environ.get('AWS_ACCESS_KEY_ID', 'DUMMY_KEY'),
        aws_secret_access_key=os.environ.get('AWS_SECRET_ACCESS_KEY', 'DUMMY_SECRET'),
        region_name=os.environ.get('AWS_REGION', 'us-east-1')
    )

    body_content = {"prompt": prompt, "max_tokens_to_sample": 200, "temperature": 0.7}
    
    # Example of type-hinted request body (mypy will check its structure)
    request_body: InvokeModelRequestRequestTypeDef = {
        'body': str(body_content).encode('utf-8'), # In real use, convert dict to JSON string
        'contentType': 'application/json',
        'accept': 'application/json',
        'modelId': model_id
    }

    try:
        response = client.invoke_model(**request_body)
        print("Successfully invoked model:")
        # Process response body (example)
        response_body = response['body'].read().decode('utf-8')
        print(response_body)
    except client.exceptions.AccessDeniedException as e:
        print(f"Access Denied: {e}")
        print("Please ensure your AWS credentials and permissions are correctly configured.")
    except Exception as e:
        print(f"An error occurred: {e}")


# Example usage:
if __name__ == "__main__":
    # Replace with an actual model ID available in your region
    # e.g., 'anthropic.claude-v2' or 'amazon.titan-text-express-v1'
    model = "DUMMY_MODEL_ID" 
    user_prompt = "Hello, how are you?"
    invoke_bedrock_model(model, user_prompt)

view raw JSON →