Docs

Base URL https://api.auxilio.ai/v1. Any OpenAI-compatible client works: set the base URL and your key.

Quickstart

curl

curl https://api.auxilio.ai/v1/chat/completions \
  -H "Authorization: Bearer $AUXILIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen/qwen3.8-27b",
    "messages": [{"role": "user", "content": "Say hello in Romanian."}],
    "stream": true
  }'

Python, OpenAI SDK

from openai import OpenAI

client = OpenAI(base_url="https://api.auxilio.ai/v1", api_key="YOUR_KEY")
r = client.chat.completions.create(
    model="qwen/qwen3.8-27b",
    messages=[{"role": "user", "content": "Say hello in Romanian."}],
)
print(r.choices[0].message.content)

Node, OpenAI SDK

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.auxilio.ai/v1", apiKey: process.env.AUXILIO_API_KEY });
const r = await client.chat.completions.create({
  model: "qwen/qwen3.8-27b",
  messages: [{ role: "user", content: "Say hello in Romanian." }],
});
console.log(r.choices[0].message.content);

Thinking

Reasoning is on by default and billed as output tokens. Switch it off per request with a chat template argument:

r = client.chat.completions.create(
    model="qwen/qwen3.8-27b",
    messages=[{"role": "user", "content": "Return only the answer: 17 * 23"}],
    extra_body={"chat_template_kwargs": {"enable_thinking": False}},
)

In raw JSON the same field sits at the top level of the request body: "chat_template_kwargs": {"enable_thinking": false}.

Tools and structured output

Pass tools and tool_choice as with OpenAI; tool calls come back parsed in message.tool_calls. response_format with a JSON schema is supported.

Models

GET /v1/models lists what your key can call. Model ids, context, limits and prices: Models and pricing.

Rate limits

Over capacity the gateway answers 429 immediately rather than queueing; wait and retry with backoff. Streams that were already open keep running. Usage is included in every streamed chunk.

Data

Prompts and completions are not stored. Request metadata and token counts are kept for billing. Zero Data Retention, Privacy Policy.

Request an API key

Keys are issued by hand within one business day. Self-serve through OpenRouter follows once the listing is live.

What you will send through the API. One or two sentences is enough.
Terms(Required)