Thomas B.

botocore.errorfactory.ValidationException: An error occurred (ValidationException) when calling the InvokeModel operation: Invocation of model ID meta.llama4-maverick-17b-instruct-v1:0 with on-demand throughput isn't supported. Retry your request with the ID or ARN of an inference profile that contains this model.

import os
import json
from pprint import pprint

import boto3


use_arize = True
if use_arize:
    import arize.otel
    from openinference.instrumentation.bedrock import BedrockInstrumentor

    arize_space_id = os.getenv("ARIZE_SPACE_ID")
    arize_api_key = os.getenv("ARIZE_API_KEY")

    tracer_provider = arize.otel.register(
        space_id=arize_space_id,
        api_key=arize_api_key,
        project_name="tom-local",
    )
    BedrockInstrumentor().instrument(tracer_provider=tracer_provider)

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


print("\n\n=== invoke_model with llama-4 ===")

prompt = b'{"prompt": "Human: Tell me about the AWS Bedrock invoke_model method in one sentence. Assistant:"}'
invoke_response = client.invoke_model(
    body=prompt,
    contentType="application/json",
    modelId="us.meta.llama4-maverick-17b-instruct-v1:0",
)
pprint(invoke_response)
invoke_body = json.loads(invoke_response.get("body").read())
pprint(invoke_body)


print("\n\n=== converse with sonnet-4.5 ===")

converse_request = {
    "modelId": "global.anthropic.claude-sonnet-4-5-20250929-v1:0",
    "messages": [
        {
            "role": "user",
            "content": [
                {
                    "text": "Tell me about the AWS Bedrock converse method in one sentence."
                }
            ],
        }
    ],
    "inferenceConfig": {"maxTokens": 100},
}
converse_response = client.converse(**converse_request)
pprint(converse_response)