Documentation

Introduction

Tuori Cloud provides a unified, OpenAI-compatible API for model inference. One endpoint and one API key can be used across supported commercial and open models.

Quickstart

1. Create an API key.

   Generate a development key and store it as TUORI_API_KEY. Never expose secret keys in browser code.

2. Install the SDK.
   pip install openaiCopy
3. Send your first request.
   PythonCopy

   from openai import OpenAI
   import os
   
   client = OpenAI(
       api_key=os.environ["TUORI_API_KEY"],
       base_url="https://api.tuori.cloud/v1",
       timeout=30.0,
       max_retries=2,
   )
   
   response = client.chat.completions.create(
       model="your-model-id",
       messages=[
           {"role": "system", "content": "You are a helpful assistant."},
           {"role": "user", "content": "Explain vector databases in one paragraph."}
       ],
   )
   print(response.choices[0].message.content)

4. Inspect usage and request IDs.

   Record token usage, latency, status code, and the response request ID for support and debugging.

Authentication

Send your secret key as a Bearer token. Use separate keys for development and production, apply least-privilege scopes, and rotate keys regularly.

Authorization: Bearer $TUORI_API_KEY
Content-Type: application/json

SDKs

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.TUORI_API_KEY,
  baseURL: "https://api.tuori.cloud/v1",
  timeout: 30_000,
  maxRetries: 2,
});

const response = await client.chat.completions.create({
  model: "your-model-id",
  messages: [{ role: "user", content: "Hello" }],
});

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

Streaming

Enable streaming for lower perceived latency. Accumulate content deltas and handle disconnects gracefully.

stream = client.chat.completions.create(
    model="your-model-id",
    messages=[{"role": "user", "content": "Write a short launch announcement."}],
    stream=True,
)

for chunk in stream:
    delta = chunk.choices[0].delta.content
    if delta:
        print(delta, end="", flush=True)

Tool Calling

Declare tools with JSON Schema, execute validated tool calls in your application, then return tool results to the model. Never execute untrusted arguments without validation.

Retries and Timeouts

Retry rate-limit and transient server errors with exponential backoff and jitter. Do not retry invalid authentication or malformed requests.

cURL Example

curl https://api.tuori.cloud/v1/chat/completions \
  -H "Authorization: Bearer $TUORI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "your-model-id",
    "messages": [{"role": "user", "content": "Hello"}],
    "temperature": 0.2
  }'

Error Codes

Rate Limits

Limits may be enforced by requests per minute, tokens per minute, concurrent requests, or budget. Keep concurrency bounded and use a queue for bursty workloads.