Skip to main content

​
Migrate from Anthropic SDK

If you are using Anthropic’s SDK directly, switching to Routor takes two changes. You swap to the OpenAI-compatible SDK format, which Routor uses as its standard interface.

​
Before and After

​
Python (Anthropic SDK → Routor)

Before: 
import anthropic

client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])

message = client.messages.create(
    model="claude-opus-4-5",
    max_tokens=1024,
    messages=[{"role": "user", "content": "explain closures in JavaScript"}]
)

print(message.content[0].text)
After: 
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["ROUTOR_API_KEY"],
    base_url="https://api.routor.ai/v1",
)

response = client.chat.completions.create(
    model="auto",
    max_tokens=1024,
    messages=[{"role": "user", "content": "explain closures in JavaScript"}]
)

print(response.choices[0].message.content)

​
Node.js (Anthropic SDK → Routor)

Before: 
import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({ apiKey: process.env.ANTHROPIC_API_KEY });

const message = await client.messages.create({
  model:      "claude-opus-4-5",
  max_tokens: 1024,
  messages:   [{ role: "user", content: prompt }],
});

console.log(message.content[0].text);
After: 
import OpenAI from "openai";

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

const response = await client.chat.completions.create({
  model:      "auto",
  max_tokens: 1024,
  messages:   [{ role: "user", content: prompt }],
});

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

​
Key Differences

The Anthropic SDK uses a different response format than the OpenAI SDK. The main changes when moving to Routor’s OpenAI-compatible format:
Anthropic SDKRoutor (OpenAI format)
message.content[0].textresponse.choices[0].message.content
client.messages.create()client.chat.completions.create()
max_tokens parammax_tokens param (same)
claude-opus-4-5 as model"auto" as model

​
If You Still Want Claude Specifically

Routor will route to Claude when the prompt complexity justifies it. If you always want Claude regardless of complexity, set a tier floor that maps to Claude models: 
response = client.chat.completions.create(
    model="auto",
    messages=[...],
    extra_body={"routor_tier_floor": "STANDARD"}
)
Or create a Routing Profile with a tier floor set permanently, so every request from that key uses at least STANDARD tier where Claude Sonnet is the primary model.

​
System Prompts

Both SDKs support system prompts, just in different formats. In the OpenAI format used by Routor: 
response = client.chat.completions.create(
    model="auto",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user",   "content": prompt}
    ]
)