ChatBedrockConverse doesn't support structured output with Bedrock Llama 3.3

[tioans]

Description

I am trying to get browser-use to work with Llama 3.3 via Bedrock, but I get an error when trying to use structured output. Structured output is necessary to make the model work in this context. It seems the issue comes from the with_structured_output method, as I can execute calls without it, but in that case, the output is not correct.

Error

"'RunnableSequence' object has no attribute 'with_structured_output'"

Steps to reproduce

Structured output

import asyncio
from dotenv import load_dotenv
from typing import Any, Dict, List

from pydantic import BaseModel, Field

from langchain_openai import AzureChatOpenAI
from langchain_ollama import ChatOllama
from langchain_aws import ChatBedrockConverse
from browser_use import Agent, Browser, BrowserConfig

class StructuredOutput(BaseModel):

    current_state: CurrentState = Field(
            description="Contains the current state details of the browser page."
        )
    
        action: List[Dict[str, Dict[str, Any]]] = Field(
            description=(
                "A list of actions to be executed in sequence. Each action should have one key "
                "that represents the action name and a value with the corresponding parameters."
            )
        )
 

Main method:

async def main():

    llm = ChatBedrockConverse(
            model_id="meta.llama3-3-70b-instruct-v1:0",
            temperature=0,
            verbose=True
        )

    structured_llm = llm.with_structured_output(StructuredOutput, include_raw=False)

    agent = Agent(
        task="Search for the weather in Copenhagen and return the latest.",
        llm=structured_llm,
        validate_output=True
    )
    result = await agent.run()
    print(result)


if __name__ == "__main__":
    asyncio.run(main())