Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/MemoriLabs/Memori/llms.txt

Use this file to discover all available pages before exploring further.

​
AWS Bedrock Integration

Memori integrates with AWS Bedrock through the LangChain ChatBedrock adapter, providing memory for Claude, Llama, Mistral, and other Bedrock-hosted models.

​
Installation

pip install memori langchain-aws boto3

​
Quick Start

from langchain_aws import ChatBedrock
from memori import Memori

client = ChatBedrock(
    model_id="anthropic.claude-sonnet-4-5-20250929",
    region_name="us-east-1"
)

# Register the Bedrock client with Memori
mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="bedrock_agent")

response = client.invoke("Hello! My name is Alice.")
print(response.content)

​
AWS Configuration

Bedrock requires AWS credentials. Configure them using environment variables or AWS CLI: 
export AWS_ACCESS_KEY_ID="your-access-key"
export AWS_SECRET_ACCESS_KEY="your-secret-key"
export AWS_DEFAULT_REGION="us-east-1"
Or pass credentials directly: 
from langchain_aws import ChatBedrock
from memori import Memori
import boto3

session = boto3.Session(
    aws_access_key_id="your-access-key",
    aws_secret_access_key="your-secret-key",
    region_name="us-east-1"
)

client = ChatBedrock(
    model_id="anthropic.claude-sonnet-4-5-20250929",
    client=session.client("bedrock-runtime")
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="bedrock_secure")

​
Available Models

Memori supports all Bedrock models through ChatBedrock:

​
Anthropic Claude

from langchain_aws import ChatBedrock
from memori import Memori

# Claude Sonnet 3.5
client = ChatBedrock(
    model_id="anthropic.claude-3-5-sonnet-20241022-v2:0",
    region_name="us-east-1"
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="claude_sonnet")

​
Meta Llama

from langchain_aws import ChatBedrock
from memori import Memori

# Llama 3.1 70B
client = ChatBedrock(
    model_id="meta.llama3-1-70b-instruct-v1:0",
    region_name="us-east-1"
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="llama")

​
Mistral

from langchain_aws import ChatBedrock
from memori import Memori

# Mistral Large
client = ChatBedrock(
    model_id="mistral.mistral-large-2407-v1:0",
    region_name="us-east-1"
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="mistral")

​
Multi-Turn Conversations

LangChain provides conversation history management: 
from langchain_aws import ChatBedrock
from langchain.schema import HumanMessage, AIMessage
from memori import Memori

client = ChatBedrock(
    model_id="anthropic.claude-sonnet-4-5-20250929",
    region_name="us-east-1"
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_456", process_id="chat")

messages = []

# First turn
messages.append(HumanMessage(content="My favorite programming language is Python."))
response = client.invoke(messages)
messages.append(AIMessage(content=response.content))

# Second turn - memory maintained
messages.append(HumanMessage(content="What's my favorite language?"))
response = client.invoke(messages)
print(response.content)

​
Model-Specific Parameters

Configure model-specific parameters through model_kwargs: 
from langchain_aws import ChatBedrock
from memori import Memori

client = ChatBedrock(
    model_id="anthropic.claude-sonnet-4-5-20250929",
    region_name="us-east-1",
    model_kwargs={
        "temperature": 0.7,
        "top_p": 0.9,
        "max_tokens": 2048
    }
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="custom_params")

response = client.invoke("Explain quantum computing")
print(response.content)

​
Tool Use / Function Calling

Bedrock models support function calling: 
from langchain_aws import ChatBedrock
from langchain.tools import tool
from memori import Memori

@tool
def get_weather(location: str) -> str:
    """Get the current weather for a location."""
    return f"Weather in {location}: Sunny, 72°F"

client = ChatBedrock(
    model_id="anthropic.claude-sonnet-4-5-20250929",
    region_name="us-east-1"
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="tools")

client_with_tools = client.bind_tools([get_weather])
response = client_with_tools.invoke("What's the weather in Seattle?")

if response.tool_calls:
    for tool_call in response.tool_calls:
        print(f"Tool: {tool_call['name']}")
        print(f"Args: {tool_call['args']}")

​
Streaming Responses

Stream responses for real-time applications: 
from langchain_aws import ChatBedrock
from memori import Memori

client = ChatBedrock(
    model_id="anthropic.claude-sonnet-4-5-20250929",
    region_name="us-east-1",
    streaming=True
)

mem = Memori().llm.register(chatbedrock=client)
mem.attribution(entity_id="user_123", process_id="streaming")

for chunk in client.stream("Write a poem about AI"):
    print(chunk.content, end="", flush=True)

​
Supported Features

FeatureSupportMethod
Sync Client✓client.invoke()
Async Client✓await client.ainvoke()
Streaming✓client.stream()
Tool Use✓bind_tools()
System Prompts✓LangChain message types
Custom Parameters✓model_kwargs
Cross-Region✓region_name parameter

​
How It Works

When you register a ChatBedrock client with Memori:
  1. Memori wraps the Bedrock runtime client’s invoke_model methods
  2. All requests (model ID, messages, parameters) are captured
  3. All responses are captured, including streaming chunks
  4. Conversations are stored in your Memori memory store
  5. A knowledge graph is built from conversation patterns
  6. Original LangChain behavior is preserved
Memori captures streaming responses by collecting chunks and reconstructing the full conversation after streaming completes.

​
Supported Model Families

ProviderModel FamilyExample Model ID
AnthropicClaudeanthropic.claude-3-5-sonnet-20241022-v2:0
MetaLlamameta.llama3-1-70b-instruct-v1:0
MistralMistralmistral.mistral-large-2407-v1:0
AmazonTitanamazon.titan-text-premier-v1:0
AI21 LabsJurassicai21.j2-ultra-v1
CohereCommandcohere.command-text-v14

​
Regional Availability

Bedrock models are available in specific AWS regions. Common regions:
  • us-east-1 (N. Virginia)
  • us-west-2 (Oregon)
  • eu-west-1 (Ireland)
  • ap-southeast-1 (Singapore)
Check AWS Bedrock documentation for current availability.

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us