If your code already uses the OpenAI Python or JavaScript SDK, switching to AIGate is one line:
import openai
client = openai.OpenAI(
api_key="sk-aigate-...",
base_url="https://api.aigatecloud.com/v1",
)That's it. Pass model="quality" instead of gpt-4o and AIGate routes you to the best Llama 3.3 70B inference (Groq) with retry + fallback. response_format, temperature, max_tokens all work as expected.
Full reference: Chat completions endpoint.