From Playground to Code: Make Your First AIOZ AI API Call
TL;DR
The AIOZ AI API brings models you explored in Playground into your own code. Create a key, choose a callable model, copy its model-specific sample, and make your first request in minutes. Start with one call. Inspect the response. Then extend the same workflow into a batch, backend, or pipeline.
What Is the AIOZ AI API?
AIOZ AI API lets you call the same models you tested in Playground from your own scripts, applications, and workflows.
Each callable model includes an API Reference tab beside the Playground. It contains:
- The endpoint
- Expected inputs
- Output schema
- Ready-to-run examples for cURL, Python, Node.js, and Go
The path from exploration to integration is intentionally short: choose a model, copy the sample, add your key, and run.
Note: API access is shipping to accounts in waves. If Settings → API Keys is not visible yet, your account is queued for an upcoming rollout wave.
Before You Begin
You will need:
- An AIOZ AI account with access to Settings → API Keys
- A callable model with an API Reference tab
- A terminal or development environment
- cURL for a quick test, or an official SDK for your preferred language
Step 1: Create an API Key
Open Settings → API Keys and create a key. Keep the key outside your source files.
A personal key works well for individual use. For team workflows, use an organization key so access belongs to the organization rather than a single account.
Store the key as an environment variable:
export AIOZ_AI_API_KEY="paste-your-key-here"
Step 2: Choose a Callable Model
Open a model page and look for the API Reference tab beside Playground.
That tab is the signal that a model is callable.
It already contains everything you need:
- Endpoint
- Input parameters
- Output schema
- Language-specific request examples
You do not need to memorize the format. The model page generates the correct sample for that model.
Step 3: Make Your First Request
Choose the example that matches your stack:
- cURL
- Python
- Node.js
- Go
For the fastest first test, start with cURL.
Copy the model-specific sample from the API Reference tab, add your key, and run it in your terminal.
The response returns as JSON.
That single request demonstrates the full API flow:
inputs in → model runs → structured output returned
For a deeper walkthrough of request and response behavior, read the API usage guide.
Step 4: Move to an Official SDK
Once your first request works, use the official SDK for your language:
The SDK handles authentication and transport so your code can focus on the inputs you send and what you do with the result
Step 5: Read the Response
The returned fields follow the output schema shown in the model's API Reference tab.
The documented fields are the fields you receive.
From there, you can:
- Pass results into an application
- Add them to a queue
- Store them in a workflow
- Feed them into another model
That is the core loop:
Key → Model → Input → Output
Where to Go Next
One successful request opens the door to broader workflows.
Try one of these next steps:
- Process datasets through a loop or batch
- Add a model call to a backend route
- Chain outputs between models to build a pipeline
- Build internal tools that use models without relying on a UI
Start Building
You have already tested the model in Playground. Now run it in your own code.
Pick a model, create a key, copy the sample, and make your first API call.
Explore callable models and start building with AIOZ AI.
FAQ
Q1: How do I know whether my account has API access?
Open Settings and look for API Keys. If the section is not visible yet, your account is waiting for an upcoming rollout wave.
Q2: How do I know whether a model is callable?
Open the model page. Callable models include an API Reference tab beside Playground.
Q3: Should I start with cURL or an SDK?
Use cURL for the fastest first request. Move to an official SDK when integrating the model into an application or workflow.