AWS Bedrock Python SDK Integration

Learn how to integrate AWS Bedrock with Helicone using Python.

This integration method is maintained but no longer actively developed. For the best experience and latest features, use our new AI Gateway with unified API access to 100+ models.

Proxy Integration

1. Create an account + Generate an API Key

   Log into helicone or create an account. Once you have an account, you can generate an API key.

2. Set HELICONE_API_KEY as an environment variable

   HELICONE_API_KEY=<your API key>
   
3. Run the Bedrock command with the Helicone Proxy

   import boto3
   import json
   import os
   client = boto3.client("bedrock-runtime", region_name="ap-south-1",
                         endpoint_url="https://bedrock.helicone.ai/v1/ap-south-1")
   model_id = "meta.llama3-8b-instruct-v1:0"
   
   event_system = client.meta.events
   
   
   def process_custom_arguments(params, context, **kwargs):
       if (custom_headers := params.pop("custom_headers", None)):
           context["custom_headers"] = custom_headers
   
   
   def add_custom_header_before_call(model, params, request_signer, **kwargs):
       params['headers']['Helicone-Auth'] = f'Bearer {os.getenv("HELICONE_API_KEY")}'
       params['headers']['aws-access-key'] = '<AWS ACCESS KEY>'
       params['headers']['aws-secret-key'] = '<AWS SECRET KEY>'
       # optionally, you can pass the aws-session-token instead of access and secret key if you are using temporary credentials
       params['headers']['aws-session-token'] = '<AWS SESSION TOKEN>'
       if (custom_headers := params.pop("custom_headers", None)):
           params['headers'].update(custom_headers)
       headers = params['headers']
       print(f'param headers: {headers}')
   
   
   event_system.register("before-parameter-build.bedrock-runtime.InvokeModel",
                         process_custom_arguments)
   event_system.register('before-call.bedrock-runtime.InvokeModel',
                         add_custom_header_before_call)
   
   
   body = {
       "prompt": "Hello, world!",
       "max_gen_len": 512,
       "temperature": 0.5,
       "top_p": 0.9
   }
   
   response = client.invoke_model(
       body=json.dumps(body),
       modelId=model_id,
       accept="application/json",
       contentType="application/json",
   )
   model_response = json.loads(response["body"].read())
   print(model_response)
   

Async Integration

1. Create an account + Generate an API Key

   Log into Helicone or create an account. Once you have an account, you can generate an API key.

2. Set API keys as environment variables

   export HELICONE_API_KEY=<your Helicone API key>
   export AWS_ACCESS_KEY_ID=<your AWS access key ID>
   export AWS_SECRET_ACCESS_KEY=<your AWS secret access key>
   export AWS_REGION=<your AWS region>
   
3. Install necessary packages

   Ensure you have the necessary packages installed in your JavaScript project:

   pip install helicone-async boto3
   
4. Import required modules and initialize Helicone Logger

   from helicone_async import HeliconeAsyncLogger
   import boto3
   import json
   import os
   
   logger = HeliconeAsyncLogger(api_key=os.getenv("HELICONE_API_KEY"))
   logger.init()
   
5. Configure AWS Bedrock client

   client = boto3.client("bedrock-runtime", region_name=os.getenv("AWS_REGION"))
   
6. Send the request and handle the response

   
   response = client.invoke_model(
       body=json.dumps(body),
       modelId=model_id,
       accept="application/json",
       contentType="application/json",
   )
   model_response = json.loads(response["body"].read())
   print(model_response)
   

This integration allows you to use AWS Bedrock with Helicone’s async logging. The Helicone logger will automatically capture traces of your AWS Bedrock API calls, providing you with valuable insights and analytics through the Helicone dashboard.

For more information on using async logging, visit the Async Logging documentation.