Back to Docs
Integration Guide

Meridian with LlamaIndex

Use Meridian as the LLM backend for LlamaIndex RAG pipelines. Swap OpenAI for Meridian with a single api_base change — no code rewrites, no vendor lock-in.

Installation

pip install llama-index llama-index-llms-openai-like

Configuration

Set your Meridian API key and base URL. Use environment variables or pass them directly.

MERIDIAN_API_KEY=sk-your-meridian-key
MERIDIAN_API_BASE=https://api.getnimbus.net/v1

Full Example

A complete RAG pipeline: load documents, build a vector index, and query with streaming.

from llama_index.llms.openai_like import OpenAILike
from llama_index.core import Settings, VectorStoreIndex, SimpleDirectoryReader

llm = OpenAILike(
    model="meridian-pro",
    api_base="https://api.getnimbus.net/v1",
    api_key="sk-your-meridian-key",
    temperature=0.7,
    max_tokens=4096,
    is_chat_model=True,
)

Settings.llm = llm
Settings.chunk_size = 1024
Settings.chunk_overlap = 128

documents = SimpleDirectoryReader("./docs").load_data()
index = VectorStoreIndex.from_documents(documents)

query_engine = index.as_query_engine(streaming=True)
response = query_engine.query("Summarize the key findings.")
response.print_response_stream()

Why Meridian + LlamaIndex?

  • Drop-in OpenAI replacement — same OpenAILike interface, zero refactoring.
  • No data leaves your pipeline — Meridian runs on your infrastructure or ours, your choice.
  • Streaming responses with print_response_stream() for real-time UX.
  • Full LlamaIndex ecosystem — agents, routers, retrievers, all backed by Meridian.
More IntegrationsGet API Key