Migrating from OpenAI

If your code uses the OpenAI SDK, migrating is two lines: point the base URL at AIx and use your AIx key. The request/response shapes are the same.

The change

from openai import OpenAI

client = OpenAI(
    base_url="https://api.aix.theaimart.co/v1",   # ← changed
    api_key="YOUR_AIX_KEY",                        # ← changed
)
# everything else is identical
client.chat.completions.create(
    model="anthropic/claude-sonnet-4-6",
    messages=[{"role": "user", "content": "hi"}],
)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.aix.theaimart.co/v1", // ← changed
  apiKey: process.env.AIX_KEY,                // ← changed
});

# Many tools read these — just point them at AIx:
export OPENAI_BASE_URL="https://api.aix.theaimart.co/v1"
export OPENAI_API_KEY="YOUR_AIX_KEY"

What’s the same

• chat.completions (streaming + non-streaming), embeddings, images.generations, audio.speech, audio.transcriptions, models.
• The OpenAI error envelope ({ "error": { "message", "type", "code" } }).
• Request params (temperature, max_tokens, top_p, stop, …) and response fields.

What changes

• Model names are canonical, provider-prefixed — e.g. anthropic/claude-sonnet-4-6, deepseek-ai/DeepSeek-V3.2, not gpt-4o. See the model catalog.
• You get automatic route failover — AIx selects eligible paths and recovers from a degraded route without changing your client contract.
• Billing is prepaid credits across all of theaimart — see Billing.

Not (yet) supported

• Assistants API, fine-tuning, the Batch API, and the Realtime API are not part of the /v1 surface today. The core chat / embeddings / images / audio endpoints are.