AWS Bedrock
Overview
Amazon Bedrock is a fully managed service that provides access to foundation models from multiple providers (Anthropic, Meta, Amazon, Mistral, Cohere) through a unified AWS API. It integrates natively with AWS IAM, VPC, CloudWatch, and S3, making it ideal for enterprise workloads requiring compliance, security controls, and AWS-native data pipelines.
Setup
pip install boto3
# Configure AWS credentials
aws configure
# Or set environment variables:
export AWS_ACCESS_KEY_ID=...
export AWS_SECRET_ACCESS_KEY=...
export AWS_DEFAULT_REGION=us-east-1
Enable model access in the AWS Console: Bedrock → Model Access → Enable models
Available Models
| Model ID | Provider | Best For |
|---|
anthropic.claude-3-5-sonnet-20241022-v2:0 | Anthropic | Best overall quality |
anthropic.claude-3-5-haiku-20241022-v1:0 | Anthropic | Fast, cost-efficient |
anthropic.claude-3-opus-20240229-v1:0 | Anthropic | Most capable reasoning |
meta.llama3-70b-instruct-v1:0 | Meta | Open-weight, Llama 3 70B |
meta.llama3-8b-instruct-v1:0 | Meta | Fast, smaller Llama |
amazon.titan-text-express-v1 | Amazon | AWS-native text generation |
mistral.mistral-large-2402-v1:0 | Mistral | Code + reasoning |
cohere.command-r-plus-v1:0 | Cohere | RAG, tool use |
Instructions
Converse API (Recommended)
The Converse API is the unified chat interface for all Bedrock models:
import boto3
bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")
response = bedrock.converse(
modelId="anthropic.claude-3-5-sonnet-20241022-v2:0",
messages=[
{"role": "user", "content": [{"text": "Explain AWS Lambda in simple terms."}]}
],
system=[{"text": "You are a helpful AWS solutions architect."}],
inferenceConfig={
"maxTokens": 1024,
"temperature": 0.7,
},
)
print(response["output"]["message"]["content"][0]["text"])
print(f"Input tokens: {response['usage']['inputTokens']}")
print(f"Output tokens: {response['usage']['outputTokens']}")
Streaming with Converse
import boto3
bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")
response = bedrock.converse_stream(
modelId="anthropic.claude-3-5-sonnet-20241022-v2:0",
messages=[{"role": "user", "content": [{"text": "Write a Python quicksort implementation."}]}],
)
for event in response["stream"]:
if "contentBlockDelta" in event:
delta = event["contentBlockDelta"]["delta"]
if "text" in delta:
print(delta["text"], end="", flush=True)
print()
InvokeModel API (Raw)
For models not yet supported by Converse, or for direct access:
import boto3
import json
bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")
# Claude via InvokeModel
body = {
"anthropic_version": "bedrock-2023-05-31",
"max_tokens": 1024,
"messages": [
{"role": "user", "content": "What is the capital of France?"}
],
}
response = bedrock.invoke_model(
modelId="anthropic.claude-3-5-sonnet-20241022-v2:0",
body=json.dumps(body),
contentType="application/json",
accept="application/json",
)
result = json.loads(response["body"].read())
print(result["content"][0]["text"])
Multi-Modal — Image Analysis
import boto3
import base64
import json
bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")
# Read image
with open("diagram.png", "rb") as f:
image_b64 = base64.b64encode(f.read()).decode()
response = bedrock.converse(
modelId="anthropic.claude-3-5-sonnet-20241022-v2:0",
messages=[
{
"role": "user",
"content": [
{
"image": {
"format": "png",
"source": {"bytes": base64.b64decode(image_b64)},
}
},
{"text": "Describe this architecture diagram and identify potential issues."},
],
}
],
)
print(response["output"]["message"]["content"][0]["text"])
Tool Use (Function Calling)
import boto3
import json
bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")
tools = [
{
"toolSpec": {
"name": "query_database",
"description": "Execute a SQL query against the production database",
"inputSchema": {
"json": {
"type": "object",
"properties": {
"sql": {"type": "string", "description": "SQL query to execute"},
"database": {"type": "string", "description": "Database name"},
},
"required": ["sql"],
}
},
}
}
]
messages = [{"role": "user", "content": [{"text": "How many active users do we have?"}]}]
response = bedrock.converse(
modelId="anthropic.claude-3-5-sonnet-20241022-v2:0",
messages=messages,
toolConfig={"tools": tools},
)
# Handle tool use
if response["stopReason"] == "tool_use":
tool_use = next(b for b in response["output"]["message"]["content"] if "toolUse" in b)
print(f"Tool: {tool_use['toolUse']['name']}")
print(f"Input: {tool_use['toolUse']['input']}")
# Return tool result
messages.append(response["output"]["message"])
messages.append({
"role": "user",
"content": [
{
"toolResult": {
"toolUseId": tool_use["toolUse"]["toolUseId"],
"content": [{"json": {"count": 12483, "active_last_30d": 8921}}],
}
}
],
})
final = bedrock.converse(
modelId="anthropic.claude-3-5-sonnet-20241022-v2:0",
messages=messages,
toolConfig={"tools": tools},
)
print(final["output"]["message"]["content"][0]["text"])
Knowledge Bases for RAG
import boto3
# Knowledge Base RAG — Bedrock manages embedding and retrieval
bedrock_agent = boto3.client("bedrock-agent-runtime", region_name="us-east-1")
# Retrieve relevant documents
retrieve_response = bedrock_agent.retrieve(
knowledgeBaseId="KB123456789",
retrievalQuery={"text": "What is our refund policy?"},
retrievalConfiguration={
"vectorSearchConfiguration": {"numberOfResults": 5}
},
)
# Extract text from results
contexts = [r["content"]["text"] for r in retrieve_response["retrievalResults"]]
# Generate answer grounded in retrieved docs
retrieve_and_generate = bedrock_agent.retrieve_and_generate(
input={"text": "What is our refund policy?"},
retrieveAndGenerateConfiguration={
"type": "KNOWLEDGE_BASE",
"knowledgeBaseConfiguration": {
"knowledgeBaseId": "KB123456789",
"modelArn": "arn:aws:bedrock:us-east-1::foundation-model/anthropic.claude-3-5-sonnet-20241022-v2:0",
},
},
)
print(retrieve_and_generate["output"]["text"])
Guardrails for Content Safety
import boto3
bedrock = boto3.client("bedrock-runtime", region_name="us-east-1")
# Apply a guardrail to filter content
response = bedrock.converse(
modelId="anthropic.claude-3-5-sonnet-20241022-v2:0",
messages=[{"role": "user", "content": [{"text": "User's message here"}]}],
guardrailConfig={
"guardrailIdentifier": "my-guardrail-id",
"guardrailVersion": "DRAFT", # or "1", "2", etc.
"trace": "enabled",
},
)
# Check if content was blocked
if response.get("trace", {}).get("guardrail", {}).get("inputAssessment"):
print("Guardrail assessment:", response["trace"]["guardrail"])
IAM Policy for Bedrock
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"bedrock:InvokeModel",
"bedrock:InvokeModelWithResponseStream",
"bedrock:Converse",
"bedrock:ConverseStream"
],
"Resource": [
"arn:aws:bedrock:us-east-1::foundation-model/anthropic.claude-3-5-sonnet-20241022-v2:0"
]
}
]
}
Guidelines
- Use the Converse API for new integrations — it's model-agnostic and handles message formatting.
- Enable models in the AWS Console before first use — they are not enabled by default.
- Bedrock processes data in the selected AWS region — choose for data residency compliance.
- Knowledge Bases handle chunking, embedding, and retrieval automatically with OpenSearch Serverless.
- Guardrails can block harmful content, PII, and off-topic queries before they reach the model.
- Use
converse_stream for user-facing features to reduce perceived latency.
- Cross-region inference profiles let you automatically fall back to other regions if capacity is unavailable.
- Monitor costs with AWS Cost Explorer; tag Bedrock calls with application-specific tags.