Memakai LLM API

Guide to creating API keys, subscribing to models, and using the OpenAI-compatible LLM API

OpenAI-Compatible API

Create an API Key

Consumer API Key

An API key is your credential to call the LLM API. Each key can subscribe to multiple models.

1. 01Log in and go to Console → LLM → My API Keys
2. 02Click the Create New Key. button. Set a memorable name (e.g. "key-dev", "key-production")
3. 03The key will be in the format sk-xxxxxxxxxxxx.Copy and store the key safely — you can only view it by clicking "Reveal"

Routing Strategy

Subscribe to a Model

Choose a model from the Marketplace

Each API key must subscribe to at least 1 model before it can be used.

1. 01Open the key detail drawer → select the "Subscriptions" tab
2. 02Click Subscribe More Models → browse the Marketplace and select a model
3. 03View pricing Token or Per-Request, check the rating, then click "Use Now"
4. 04(Optional) Set an Alias for the subscription — groups multiple models under a single name for routing

Call the API

Endpoint & Authentication

View the endpoint URL on the LLM Marketplace → "API Info" button. Use the API key in the header Authorization: Bearer sk-xxx.

curl — Chat Completions

curl -X POST https://api.arzastore.com/v1/chat/completions  -H "Authorization: Bearer sk-your-api-key"  -H "Content-Type: application/json"  -d '{
 "model": "gpt-4o",
 "messages": [
 {"role": "user", "content": "Xin chào!"}
 ],
 "stream": false
 }'

Python — OpenAI SDK

from openai import OpenAI
client = OpenAI(
 api_key="sk-your-api-key",
 base_url="https://api.arzastore.com/v1"
)
response = client.chat.completions.create(
 model="gpt-4o", # hoặc model_id / alias bạn đã subscribe
 messages=[
 {"role": "user", "content": "Xin chào!"}
 ],
 stream=True # hỗ trợ streaming
)
for chunk in response:
 if chunk.choices[0].delta.content:
 print(chunk.choices[0].delta.content, end="")

Node.js — OpenAI SDK

import OpenAI from "openai";
const client = new OpenAI({
 apiKey: "sk-your-api-key",
 baseURL: "https://api.arzastore.com/v1",
});
const response = await client.chat.completions.create({
 model: "gpt-4o", // hoặc model_id / alias bạn đã subscribe
 messages: [
 { role: "user", content: "Xin chào!" }
 ],
 stream: true,
});
for await (const chunk of response) {
 process.stdout.write(chunk.choices[0]?.delta?.content || "");
}

Streaming

Use with Claude Code

Automatic Setup

ARZASTORE AI supports automatic Claude Code setup. Just run a single command in the terminal.

1. 01Go to Console → LLM → API Keys, open the key detail drawer
2. 02Select the "Claude Code" tab, choose your OS and map models for 3 tiers: Haiku, Sonnet, Opus
3. 03Click "Get Setup Command" → copy the one-line command into your terminal

macOS / Linux

curl -sL "https://arzastore.com/api/v1/llm/setup-claudecode?key=sk-xxx&os=linux&haiku=model1&sonnet=model2&opus=model3&fable=model4" | sh

Windows (PowerShell)

irm "https://arzastore.com/api/v1/llm/setup-claudecode?key=sk-xxx&os=windows&haiku=model1&sonnet=model2&opus=model3&fable=model4" | iex

Manual Setup

Uninstall

To remove the ARZASTORE AI configuration from Claude Code, run one of the following commands:

macOS / Linux

curl -sL "https://arzastore.com/api/v1/llm/uninstall-claudecode?os=linux" | sh

Windows (PowerShell)

irm "https://arzastore.com/api/v1/llm/uninstall-claudecode?os=windows" | iex

After Uninstalling

Supported API Endpoints

Chat completion — compatible with OpenAI, Claude, Gemini

Text completion (legacy format)

Text embeddings for vector search, RAG

List all models you have subscribed to

💡 Tips for Effective Use

• •Use aliases to group multiple providers for the same model → automatic failover when one provider fails
• •Test your key before using in production: open drawer → "Test" tab → send a test message
• •Monitor costs via the "Model Statistics" tab in the key drawer
• •Ensure sufficient balance before using — each request deducts in real time
• •Rate models after use to help the community choose better options