Guides

Quickstart

From zero to your first completion in about five minutes. unoblox speaks the OpenAI Chat Completions API, so any OpenAI SDK works — you change the base URL and the key, nothing else.

Before you begin

You need three things:

	What	Where
1	A unoblox account and workspace	Sign in to the developer portal.
2	An API key (ub-gw-…)	Create one under API Keys.
3	A prepaid ₹ balance	Top up on the Credits page — requests draw from it per token.

1. Set your API key

Keys are workspace-scoped and look like ub-gw-…. The secret is shown once at creation. Export it so the samples below can read it from the environment — never hard-code a key in source:

shell

export UNOBLOX_API_KEY="ub-gw-user_xxx-xxxxxxxxxxxxxxxx"

2. Install an SDK (optional)

curl needs nothing. For the SDK samples, install the official OpenAI client — the same one you would use against OpenAI:

install

# Python
pip install openai

# Node.js
npm install openai

3. Send your first request

A model id is an author/model slug. The samples below use <model-id> as a placeholder — swap it for a real one from the catalog.

curl

curl https://api.unoblox.ai/v1/chat/completions \
  -H "Authorization: Bearer $UNOBLOX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<model-id>",
    "messages": [
      { "role": "system", "content": "You are a concise assistant." },
      { "role": "user", "content": "Namaste! Give me one fact about UPI." }
    ]
  }'

python

import os
from openai import OpenAI

# unoblox is OpenAI-compatible — point the OpenAI SDK at the gateway.
client = OpenAI(
    base_url="https://api.unoblox.ai/v1",
    api_key=os.environ["UNOBLOX_API_KEY"],  # ub-gw-...
)

resp = client.chat.completions.create(
    model="<model-id>",
    messages=[
        {"role": "system", "content": "You are a concise assistant."},
        {"role": "user", "content": "Namaste! Give me one fact about UPI."},
    ],
)

print(resp.choices[0].message.content)

node.js

import OpenAI from "openai";

// unoblox is OpenAI-compatible — point the OpenAI SDK at the gateway.
const client = new OpenAI({
  baseURL: "https://api.unoblox.ai/v1",
  apiKey: process.env.UNOBLOX_API_KEY, // ub-gw-...
});

const resp = await client.chat.completions.create({
  model: "<model-id>",
  messages: [
    { role: "system", content: "You are a concise assistant." },
    { role: "user", content: "Namaste! Give me one fact about UPI." },
  ],
});

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

4. Stream the response

Set stream: true for token-by-token text/event-stream output. Add curl's -N flag to disable buffering:

curl · streaming

curl -N https://api.unoblox.ai/v1/chat/completions \
  -H "Authorization: Bearer $UNOBLOX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<model-id>",
    "messages": [{ "role": "user", "content": "Count to five." }],
    "stream": true
  }'

The stream is OpenAI-compatible and ends with a data: [DONE] sentinel. See Streaming for parsing details.

5. Let unoblox pick the model

Not sure which model to use? Send unoblox/auto and the router selects the best-value model that can serve the request, then fails over across providers automatically. See Routing.

curl · auto-routing

# Let unoblox pick the model for you — best value across the catalog.
curl https://api.unoblox.ai/v1/chat/completions \
  -H "Authorization: Bearer $UNOBLOX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "unoblox/auto",
    "messages": [{ "role": "user", "content": "Draft a 1-line release note." }]
  }'

If you can call OpenAI, you can call unoblox.

The request and response schemas match, so existing tools — the OpenAI SDKs, LangChain, LlamaIndex — work by changing only the base URL and key. Migrating an existing app? See Migration.

PreviousGetting started NextAuthentication