Skip to main content
Welcome! In this guide, you’ll learn how to navigate the OpenAI Platform—exploring the Playground, Dashboard, and Documentation—to streamline your AI development workflow.

​
Playground

The Playground offers an interactive environment to test models and fine-tune parameters before integrating them into your code.

​
1. Interacting with Models

Switch to the Chat tab for conversational prompts. For example: 
what is 20 times 10 plus 5 minus 3 times 10
Click Run and watch the model break down the calculation step by step, returning the answer.

​
2. Tuning Parameters

Adjust these settings to influence output quality and creativity:
  • Temperature: Controls randomness (0.0–1.0)
  • Max Tokens: Limits response length
  • Top K / Top P: Filters token selection based on probability
The image shows a chat interface in a coding environment where a mathematical expression is being solved step-by-step, resulting in the final answer of 175.

​
3. Additional Features

  • Model Switching: Try GPT-3.5, GPT-4, or custom fine-tuned models
  • Speech to Text: Real-time audio transcription
  • Text to Speech: Generate audio responses
  • Assistant Builder: Design your own conversational agent
  • Completion Settings: Preconfigure defaults for repeated experiments
The Playground is ideal for rapid prototyping. Once satisfied, copy code snippets directly into your application.

​
Dashboard

The Dashboard centralizes your account and usage management.
FeatureDescription
Usage AnalyticsView API call counts, token usage, and cost estimates
Assistants & JobsManage chat assistants and batch processing tasks
Fine-tuningUpload training data and launch custom model tweaks
File StorageStore and retrieve files for embedding or training
API KeysCreate, view, and revoke keys for secure access
Access your metrics, configure billing alerts, and optimize resource allocation all from one place.

​
Documentation

Comprehensive docs cover every aspect of the API and its components. Visit OpenAI API Documentation for full details.

​
Models Overview

Browse the Models section to compare capabilities and select the best fit:
The image shows a webpage from OpenAI's documentation, detailing an overview of various AI models, their descriptions, and capabilities. The left sidebar includes navigation options like "Overview," "Quickstart," and "Models."
Selecting a model—such as GPT-4—reveals its parameters, use cases, and any available variants.

​
Changelog

Track updates, deprecations, and new features in the Changelog. Staying informed ensures your integrations remain stable.
The image shows a changelog from a software documentation page, detailing updates and features released in August 2024, including API support and model updates.

​
Capabilities

ModalityDescription
Text GenerationConversational agents, content creation
Image GenerationDALL·E image creation and editing
AudioText-to-Speech & real-time Speech-to-Text
EmbeddingsSemantic vectors for search & clustering
VideoVideo generation and analysis (beta)

​
JavaScript Example

import OpenAI from "openai";

const openai = new OpenAI();

const completion = await openai.chat.completions.create({
  model: "gpt-4",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "Write a haiku about recursion in programming." }
  ]
});

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

​
Python Example

from openai import OpenAI

client = OpenAI()

completion = client.chat.completions.create(
    model="gpt-4",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Write a haiku about recursion in programming."}
    ]
)

print(completion.choices[0].message.content)
Be sure to set your OPENAI_API_KEY in environment variables before running these snippets.

​
Guides and Advanced Topics

Explore specialized tutorials to maximize the API’s potential:
  • Function Calling
  • Model Evaluation & Benchmarks
  • Fine-tuning & Distillation
  • Batch Processing Workflows
  • Official Libraries & SDKs
  • Prompt Engineering Best Practices
  • Rate Limits, Error Codes, and Retirements
Example: Create image variations with DALL·E in Python: 
response = client.images.create_variation(
    model="dall-e-2",
    image=open("corgi_and_cat_paw.png", "rb"),
    n=1,
    size="1024x1024"
)
print(response.data[0].url)

​
Deprecations

Regularly review the Deprecations section to avoid using sunset features and to plan migrations.
The image shows a webpage detailing deprecations and updates for various GPT models and APIs, including shutdown dates and recommended replacements. It includes a sidebar with navigation options related to best practices and resources.
Deprecated endpoints and models will stop functioning after their scheduled sunset dates. Always update to the recommended replacements to prevent downtime.

Watch Video