Langchain-aws Reasoning effort

[sriramsriram3]

Checked other resources

• This is a feature request, not a bug report or usage question.
• I added a clear and descriptive title that summarizes the feature request.
• I used the GitHub search to find a similar feature request and didn't find it.
• I checked the LangChain documentation and API reference to see if this feature already exists.
• This is not related to the langchain-community package.

Package (Required)

• langchain
• langchain-openai
• langchain-anthropic
• langchain-classic
• langchain-core
• langchain-cli
• langchain-model-profiles
• langchain-tests
• langchain-text-splitters
• langchain-chroma
• langchain-deepseek
• langchain-exa
• langchain-fireworks
• langchain-groq
• langchain-huggingface
• langchain-mistralai
• langchain-nomic
• langchain-ollama
• langchain-perplexity
• langchain-prompty
• langchain-qdrant
• langchain-xai
• Other / not sure / general

Feature Description

In Langchain-aws, its better accept the reasoning effort. Because currently, all the models are coming up with reasoning.

Use Case

I'm using the GPT model from AWS Bedrock. If the langchain-aws supported the reasoning_effort parameter (low, medium, high), it would reduce latency by a few seconds. This would be much more efficient than calling the model through raw boto3.
Currently we are unable to do it so, thats why we moved to boto3, if it provides we can back to langchain-aws

Proposed Solution

llm = ChatBedrockConverse(
model_id="gpt--",
reasoning_effort="low", # NEW PARAM
)

Alternatives Considered

No response

Additional Context

No response

[thromel]

Thanks for the feature request! However, langchain-aws is maintained in a separate repository: https://github.com/langchain-ai/langchain-aws

Please file this feature request there so the maintainers of that package can review it.

You can create a new issue at: https://github.com/langchain-ai/langchain-aws/issues/new

[thromel]

I've created a PR to address this feature request in the langchain-aws repository: #787

The PR adds a reasoning_effort parameter to ChatBedrockConverse that accepts "low", "medium", or "high" values and automatically configures the appropriate provider-specific reasoning parameters:

from langchain_aws import ChatBedrockConverse

# Amazon Nova with reasoning
llm = ChatBedrockConverse(
    model="us.amazon.nova-lite-v1:0",
    reasoning_effort="medium",
)

# OpenAI on Bedrock with reasoning
llm = ChatBedrockConverse(
    model="openai.gpt-4o-mini-2024-07-18-v1:0",
    reasoning_effort="high",
)