AI Framework

LangChain

Meridian LLM adapter for LangChain that exposes the full Meridian model router, including fallback chains and structured outputs, as a drop-in ChatModel class.

Overview

Use Meridian as a drop-in LLM replacement within your existing LangChain agentic workflows. Meridian acts as a single `ChatModel` that handles fallback routing invisibly behind the scenes.

Quickstart

Import the Meridian integration from the LangChain community package.

python
1
from langchain_community.chat_models import ChatMeridian
2
from langchain_core.messages import HumanMessage
3
chat = ChatMeridian(api_key="sk_live_...",routing_strategy="latency_optimized")chat.invoke([HumanMessage(content="Initialize sequence.")])

Installation

npm

$npm install @meridian/sdk

View Source Code

Report an Issue

DEPLOY / MERIDIAN

Start building with Meridian.

Two ways in — pick your path.

For humans

Get started with our Next.js and Python SDKs in seconds.

Start Building

Read Framework Guides

For agents

Tools to connect your autonomous agents to our edge infrastructure.

$npm install @synapse/sdk