Bedrock Central

Amazon Bedrock gives you access to hundreds of models from a single service.

Hundredsof models30+AWS RegionsPay-as-you-goYour data stays under your account's controls

Before you can call any model, Bedrock needs to know who you are.

In this tutorial we'll use an API key. You can generate one directly in the console and start calling models in under a minute. No AWS CLI setup or IAM configuration needed.

Generate an API key

Sign in to the AWS Management Console and open the Amazon Bedrock console
In the left navigation panel, select API keys
Choose either Generate short-term API key or Generate long-term API key
For long-term keys, set your desired expiration time
Choose Generate and copy your API key

Generating an API key in the Amazon Bedrock console

Set the key in your environment

export AWS_BEARER_TOKEN_BEDROCK="bedrock-api-key-..."

Call Bedrock

boto3 detects the bearer token automatically. No extra config needed.

python
import boto3, json

bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")

resp = bedrock.invoke_model(
    modelId="global.anthropic.claude-sonnet-4-6",
    body=json.dumps({
        "messages": [{"role": "user", "content": "Hello, Bedrock!"}],
        "max_tokens": 512,
    }),
)

print(json.loads(resp["body"].read()))

You can use Bedrock API keys to quickly test your setup.

For setups that require a higher level of security, we recommend using temporary IAM credentials. see security best practices in IAM for more details.

Find your model's ID and available Regions in the model cards, then call it.

Choose a model

Claude Sonnet 4.6anthropic.claude-sonnet-4-6

Choose a Region to call

US East (N. Virginia)us-east-1

Choose how requests route

Amazon Bedrock selects the optimal Region within a geography (US, EU, JP, AU).

us-east-1us-east-2us-west-1us-west-2ca-central-1ca-west-1

python
import boto3, json

bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")

resp = bedrock.invoke_model(
    modelId="us.anthropic.claude-sonnet-4-6",
    body=json.dumps({
        "messages": [{"role": "user",
            "content": "Hello, Bedrock!"}],
        "max_tokens": 512,
    }),
)

print(json.loads(resp["body"].read()))

Bedrock offers multiple APIs to fit your needs. See APIs supported by Bedrock for more details

Add capabilities like guardrails, caching, and streaming to your API calls.

python
import boto3, json

bedrock = boto3.client("bedrock-runtime")

resp = bedrock.invoke_model(
    modelId="anthropic.claude-sonnet-4-6",
    body=json.dumps({
        "anthropic_version": "bedrock-2023-05-31",
        "max_tokens": 512,
        "messages": [{"role": "user", "content": "Hello, Bedrock!"}],
    }),
)

print(json.loads(resp["body"].read()))

Go from API calls to production applications.Build and deploy agents, bring your own models, and ground them in your data.

Explore Amazon Bedrock

Get Started

Generate an API key

Set the key in your environment

Call Bedrock

Nice work! You've got the fundamentals.