Introduction to OpenAI API & ChatGPT API for Developers

Clarify key buzzwords and define ai models, then examine training data, time, GPUs, cost, and tokens, while noting hallucinations and how ChatGPT uses tools, plugins, and MCP servers.

Explore AI, machine learning, deep learning, and GenAI as nested circles, with neural networks powering deep learning and generative capabilities like translation.

Explain what GPT stands for in ChatGPT and how a pre-trained and fine-tuned generative transformer processes natural language to generate text.

Explain AI model limitations, including hallucinations, training cutoffs, and bias, showing why plausible outputs can be wrong and how researchers reduce these issues over time.

learn how to install postman, understand freemium pricing, choose browser or standalone app, create an account, set up workspaces, and apply a dark theme.

Create and verify your OpenAI account to access the ChatGPT and OpenAI APIs, including signing up, email confirmation, and phone verification.

Learn to call the chat completion endpoint of the ChatGPT API using a POST request, supply model and messages in JSON, and troubleshoot errors to obtain a 200 OK response.

Store your OpenAI API key in a Postman collection, configure bearer token authorization, and test the get models endpoint to verify GPT-3.5-turbo availability in JSON.

Learn how to obtain an OpenAI API key, view available models, and build a basic request to the ChatGPT API by logging into platform.openai.com and generating a secret key.

Learn how completions drive AI text generation across tasks, including chat, general text, and code completions, with examples like blog titles, translations, and summaries.

Explore the OpenAI Playground to interact with chat and complete models like GPT-3.5 Turbo and GPT-4. Load presets and explore examples to learn how to experiment with OpenAI APIs.

Learn how gpt-3.5-turbo powers chat completions and versatile tasks like email writing, translating text, JavaScript coding, and summarization. Configure interactions with system, user, and assistant messages for effective API use.

Explore GPT-4 versus GPT-3.5, including its chat and completion capabilities, API access, billing differences, and how to swap in Postman.

Explore text completions with the text davinci 003 model, including prompts for titles, translations, and summaries. Practice calling the completions endpoint via Postman and simulate conversations.

Compare gpt-3.5-turbo and text-davinci-003; turbo is cheaper and a good starting point, while davinci-003 offers fine-tuning and traditional completions.

Learn to generate images with DALL-E by posting prompts to the Images API, specifying the prompt, the N parameter, and the size to obtain multiple results.

Design prompts to coax high-quality outputs from OpenAI models by being explicit, detailed, and example-driven. Learn how prompt length, temperature, and non-determinism affect results and why iteration matters.

Discover how temperature controls model creativity and output variability in the OpenAI API and ChatGPT API for developers, from zero to highly creative values, with practical Playground and API examples.

Discover how OpenAI pricing uses tokens, covering input and output tokens with a GPT-3.5 turbo example, and learn to control cost by adjusting max tokens within a 4,000 token limit.

Learn to check the OpenAI API status on the status page for availability issues, including timeouts and 500 errors, and review related services like the playground and ChatGPT API.

Discover how gpt-3.5-turbo uses system, user, and assistant roles to structure prompts and set context. Learn practical tips for configuring prompts, context, and conversation history to guide the assistant's behavior.

Learn to instruct a model with examples to build a pizzeria chatbot using system messages and conversation history, guiding GPT-3.5 Turbo through a fixed menu and Postman chat completions.

Showcases generating a summary from large text using a system message and a multi-line prompt with gpt-3.5-turbo. Learn to convert multi-line text to JSON with a pre-request script.

Learn how to limit chat completions by setting max tokens, understand token counts (prompt vs completion), and balance prompts with practical examples for efficient responses.

Learn to configure the stop property to halt model generation when a token is detected, with examples using tokens like 4. and newline in Postman.

Learn how ChatGPT Work turns AI from a chat tool into a task-performing agent that reads and writes files, edits documents, and automates work across browser and desktop apps.

Install and sign in to the ChatGPT desktop app for macOS or Windows, and explore ChatGPT Work and Codex modes for knowledge work and software development.

Leverage Codex plan mode to redesign a website into a polished, readable single-page static site by analyzing HTML and generating a cleaner index.html in the project folder.

Learn to use Codex to understand and modify an existing public game repository, then implement continuous firing by removing the weapon cooldown in asteroids.js.

Learn to build automated inbox workflows with Codex using plugins, creating daily brief summaries from Outlook emails and scheduling prompts to run every weekday at 7 a.m.

Understand how agent skills enforce consistent behavior by packaging prompts and optional scripts, loaded on demand, invoked via forward slash, and how Codex skills extend capabilities like Outlook integration.

Discover how Codex's Chrome plugin connects to Google Chrome to automate browser tasks, use steer mode, and export flight search results to Excel cloud.

Explore practical tips for using the Chrome plugin with Codex to automate daily web tasks, optimize prompts, manage permissions, and keep tabs organized.

Codex can control your entire computer across apps by enabling computer use, using accessibility features and screenshots, with parallel sessions and app-specific prompts like Chess or Numbers.