Migration Guide

What changes

import OpenAI from 'openai';

const client = new OpenAI({
- baseURL: 'https://api.openai.com/v1',
- apiKey: process.env.OPENAI_API_KEY,
+ baseURL: 'https://api.aniron.ai/v1',
+ apiKey: process.env.ANIRON_API_KEY,
});

const completion = await client.chat.completions.create({
- model: 'gpt-4o',
+ model: 'aniron/gpt-4o',
  messages: [{ role: 'user', content: 'Hello!' }],
});

• base_url → https://api.aniron.ai/v1
• api_key → your Aniron key (see Authentication)
• model → prefix the existing model name with aniron/, or switch to any of the 20+ models in the catalog

What doesn't change

Everything downstream of the client construction is unchanged: request parameters (messages, temperature, max_tokens, and so on), the streaming chunk format, tool calling request/response shapes, and the error response format. If your code accesses completion.choices[0].message.content, it keeps working with no further changes.

One addition worth making

OpenAI's API has no equivalent to Aniron's fallbacks parameter. It's optional, but worth adding during migration: an ordered array of backup model IDs that Aniron retries automatically on a rate limit or server error, at no code change beyond the request body. See Model Fallbacks.

Rollout

Test the new base URL and key against a single non-production request first. Because the request and response shapes are identical, there's no client-side parsing to validate — only that the key is active and has credit. See Managing Credits before switching production traffic over.