OpenRouter: 300개 이상 모델을 연결하는 통합 LLM API 게이트웨이, 40% 비용 절감 — 2026년 설정 가이드

# API 키를 안전하게 저장
export OPENROUTER_API_KEY="sk-or-v1-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"

# 기본 채팅 완성 요청
curl -s https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-sonnet-4.5",
    "messages": [
      {"role": "user", "content": "양자 컴퓨팅을 3문장으로 설명해줘"}
    ]
  }'

# 특수 SDK 불필요 — OpenAI 클라이언트 사용
pip install openai>=1.30.0

# openrouter_demo.py
from openai import OpenAI
import os

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=os.environ.get("OPENROUTER_API_KEY"),
)

# OpenRouter를 통해 Claude Sonnet 4.5 호출
response = client.chat.completions.create(
    model="anthropic/claude-sonnet-4.5",
    messages=[
        {"role": "system", "content": "You are a helpful coding assistant."},
        {"role": "user", "content": "Write a Python function to flatten a nested list"}
    ],
    temperature=0.7,
    max_tokens=500,
)

print(response.choices[0].message.content)
print(f"사용된 모델: {response.model}")
print(f"토큰 수: {response.usage.total_tokens}")

python openrouter_demo.py

npm install openai

// openrouter-demo.ts
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://openrouter.ai/api/v1",
  apiKey: process.env.OPENROUTER_API_KEY,
});

async function main() {
  const response = await client.chat.completions.create({
    model: "openai/gpt-5",
    messages: [
      { role: "user", content: "Write a React useDebounce hook" },
    ],
  });

  console.log(response.choices[0].message.content);
}

main();

# 300개 이상의 모델과 가격 책정 나열
curl -s https://openrouter.ai/api/v1/models \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" | \
  jq '.data[] | {id: .id, pricing: .pricing}' | head -50

# openrouter_langchain.py
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate

llm = ChatOpenAI(
    model_name="anthropic/claude-sonnet-4.5",
    openai_api_key=os.environ.get("OPENROUTER_API_KEY"),
    openai_api_base="https://openrouter.ai/api/v1",
    temperature=0.7,
)

prompt = ChatPromptTemplate.from_messages([
    ("system", "You are an expert Python developer."),
    ("human", "{input}"),
])

chain = prompt | llm

result = chain.invoke({"input": "Write a FastAPI middleware for rate limiting"})
print(result.content)

# openrouter_llamaindex.py
from llama_index.llms.openai import OpenAI as LlamaOpenAI
from llama_index.core import Settings

llm = LlamaOpenAI(
    model="meta-llama/llama-4-maverick",
    api_key=os.environ.get("OPENROUTER_API_KEY"),
    api_base="https://openrouter.ai/api/v1",
    temperature=0.3,
)

Settings.llm = llm

from llama_index.core import VectorStoreIndex, Document

docs = [Document(text="OpenRouter simplifies multi-provider LLM access.")]
index = VectorStoreIndex.from_documents(docs)
query_engine = index.as_query_engine()
response = query_engine.query("What does OpenRouter do?")
print(response)

// app/api/chat/route.ts
import { createOpenRouter } from "@openrouter/ai-sdk-provider";
import { convertToModelMessages, streamText } from "ai";

const openrouter = createOpenRouter({
  apiKey: process.env.OPENROUTER_API_KEY,
});

export async function POST(req: Request) {
  const { messages } = await req.json();

  const result = streamText({
    model: openrouter("anthropic/claude-sonnet-4.5"),
    messages: await convertToModelMessages(messages),
    system: "You are a helpful assistant.",
  });

  return result.toDataStreamResponse();
}

// openrouter_demo.go
package main

import (
	"context"
	"fmt"
	"os"

	"github.com/openai/openai-go"
	"github.com/openai/openai-go/option"
)

func main() {
	client := openai.NewClient(
		option.WithBaseURL("https://openrouter.ai/api/v1"),
		option.WithAPIKey(os.Getenv("OPENROUTER_API_KEY")),
	)

	resp, err := client.Chat.Completions.New(context.Background(), openai.ChatCompletionNewParams{
		Model: openai.String("google/gemini-3-pro"),
		Messages: openai.F([]openai.ChatCompletionMessageParamUnion{
			openai.UserMessage("Explain Go concurrency patterns"),
		}),
	})
	if err != nil {
		panic(err)
	}

	fmt.Println(resp.Choices[0].Message.Content)
}

# OpenRouter가 자동으로 최적의 모델 선택
response = client.chat.completions.create(
    model="openrouter/auto",  # 58개 이상의 후보 모델에서 자동 선택
    messages=[
        {"role": "user", "content": "Write a Kubernetes deployment YAML"}
    ],
    extra_body={
        "provider": {
            "sort": "price",
        }
    }
)
print(response.model)

# 프로덕션 장애 조치 설정
response = client.chat.completions.create(
    model="anthropic/claude-sonnet-4.5",
    messages=[{"role": "user", "content": "Critical business analysis..."}],
    extra_body={
        "provider": {
            "order": ["Anthropic", "OpenAI", "Google"],
            "allow_fallbacks": True,
        }
    }
)

# 직접 제공업체 키 저장
curl -X POST https://openrouter.ai/api/v1/credentials \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "openai",
    "key": "sk-proj-your-direct-openai-key"
  }'

# 가장 저렴한 사용 가능한 모델로 라우팅
response = client.chat.completions.create(
    model="openrouter/auto",
    messages=[{"role": "user", "content": "Summarize this article"}],
    extra_body={
        "provider": {
            "sort": "price",
            "quantizations": ["fp8", "fp16"],
        }
    }
)

# Dockerfile.openrouter-proxy
FROM node:20-alpine
WORKDIR /app
COPY package*.json ./
RUN npm install express axios
COPY . .
EXPOSE 3000
CMD ["node", "proxy.js"]

# docker-compose.yml
version: "3.8"
services:
  openrouter-proxy:
    build:
      context: .
      dockerfile: Dockerfile.openrouter-proxy
    ports:
      - "3000:3000"
    environment:
      - OPENROUTER_API_KEY=${OPENROUTER_API_KEY}
      - FALLBACK_MODELS=openai/gpt-5,google/gemini-3-pro
      - CACHE_ENABLED=true
    restart: unless-stopped

# 사용량과 비용을 프로그래밍 방식으로 추적
import requests

headers = {"Authorization": f"Bearer {os.environ.get('OPENROUTER_API_KEY')}"}

usage = requests.get(
    "https://openrouter.ai/api/v1/credits",
    headers=headers
).json()

print(f"남은 크레딧: ${usage['data']['total_credits'] - usage['data']['total_usage']}")
print(f"총 사용량: ${usage['data']['total_usage']}")

model = "anthropic/claude-sonnet-4.5"  # 또는 "openai/gpt-5" 또는 "google/gemini-3-pro"
response = client.chat.completions.create(
    model=model,
    messages=[{"role": "user", "content": "Hello!"}]
)