Core ChatGPT Features

ChatGPT is not just a chat window. It is a work environment that combines search, file analysis, images, voice, document editing, task scheduling, company data connections, research, and coding support. Instead of memorizing feature names, it is easier to choose by asking what job you are trying to do.

This page reflects official information checked on June 14, 2026. Feature names, limits, and plan availability can change, so check the OpenAI Help Center for current details.

Feature Map

Feature	Main use
Chat	Questions, writing, planning, explanation, idea organization
Web search	Current information, public sources, citation checks
File upload	Reading PDFs, tables, images, and documents
Data analysis	Summarizing, filtering, charting, and analyzing structured data
Image generation and image understanding	Creating images, describing images, exploring visual ideas
Voice mode	Spoken conversation, mobile use, speaking practice
Canvas	Editing a document or code artifact next to the chat
Projects	Managing related chats, files, and instructions
Tasks	Reminders and scheduled follow-ups
Custom GPTs	Creating a purpose-specific ChatGPT experience
Apps / Connectors	Connecting tools such as Google Drive, SharePoint, GitHub, and Slack
Deep Research	Producing structured research reports across sources
ChatGPT agent	Completing complex tasks that combine web actions, research, files, and spreadsheet editing
Codex	Writing, fixing, reviewing, and shipping code
Sora	Video generation and video-related creative workflows

Chat

Chat is the core ChatGPT experience. Use it for questions, writing, planning, code explanation, learning checks, and brainstorming.

For beginners, the key is not to write a perfect prompt on the first try. Add the goal, background, constraints, and output format step by step to improve the answer.

Web Search

Web search lets ChatGPT reference public web information. Use it for product details, news, documentation, pricing, release notes, and anything likely to change.

Common uses:

• Checking current official documentation
• Comparing products or pricing
• Creating a sourced research note
• Verifying that an answer is not based on stale knowledge

File Upload

File upload lets you ask questions about PDFs, text files, images, spreadsheets, and similar materials. Use it to summarize documents, extract arguments, compare drafts, or explain image contents.

If files include confidential information, consider whether an organizational workspace such as Business, Enterprise, or Edu is required under your data policy.

Data Analysis

Data analysis helps read, aggregate, filter, chart, and interpret structured data. It is especially useful for CSV and Excel workflows.

Good fits:

• Monthly sales summaries
• Survey categorization
• Outlier checks
• Tables and charts
• Explaining analysis steps

Image Generation and Image Understanding

ChatGPT can generate images and explain image contents. Use it for rough blog visuals, UI direction, learning diagrams, and image inspection.

For brand logos, people, copyrighted material, or high-stakes domains such as medical, legal, and financial work, review generated output carefully before using it.

Voice Mode

Voice mode lets you talk with ChatGPT. It works well on mobile, during travel, for speaking practice, for ideation, or when discussing something on screen.

Available voices, video, screensharing, and usage limits vary by plan and environment. Voice mode may not support every tool, so use the normal chat interface for file-heavy or image-generation workflows.

Canvas

Canvas opens a document or code artifact beside the chat so you can revise it repeatedly. It is useful for long writing, specs, email drafts, code, and design notes.

If chat is the conversation history, Canvas is the working draft. Use it when the output itself needs iteration.

Projects

Projects group related chats, files, and instructions. They let you keep context for an ongoing project without restating the same background every time.

Examples:

• Planning a new service
• Writing hiring or marketing materials
• Organizing research literature
• Drafting product specifications

Tasks

Tasks let ChatGPT remind you or follow up at a scheduled time. Use them for morning planning, weekly reviews, periodic checks, and deadline reminders.

They are a useful first step into automation. For workflows that actually operate external systems, consider how Tasks relate to Apps, Codex, APIs, and internal tools.

Custom GPTs

Custom GPTs are purpose-specific ChatGPT configurations with custom behavior, knowledge, and tools. They can standardize support workflows, writing review, learning help, or internal policy checks.

In Business and Enterprise, workspace GPTs can be shared and managed inside the organization. Pay close attention to sharing scope and referenced data.

Apps / Connectors

Apps and Connectors let ChatGPT access external tools and internal information. OpenAI has described a move toward using Apps as the unified term for both interactive apps and earlier connectors.

Common connections include Google Drive, SharePoint, Slack, GitHub, and Atlassian. In organizations, admins should manage enablement, permissions, roles, and action controls.

Deep Research

Deep Research is for producing structured research reports across multiple sources. It is better suited than normal search for longer investigations, comparisons, background summaries, and sourced decision memos.

Good fits:

• Market research
• Technology comparisons
• Initial review of laws, regulations, or standards
• Competitor analysis
• Long decision documents

ChatGPT Agent

ChatGPT agent handles multi-step online work by combining research with website interaction, uploaded-file processing, information from Apps, form completion, and spreadsheet editing. OpenAI introduced it on July 17, 2025. On supported plans, start it by selecting agent mode from the tools menu or typing /agent in the composer.

Depending on the task, ChatGPT agent can use a visual browser, Code Interpreter for code and data analysis, Apps, and a terminal for supported commands. You can provide more instructions while it works, take over the browser, or stop the task.

Good fits:

• Researching several websites and compiling the findings into a table or document
• Reviewing calendar or email information to organize a schedule
• Completing forms and other structured online workflows
• Updating an existing spreadsheet with analysis results

When ChatGPT agent uses signed-in websites or Apps, it may access sensitive information such as email, files, and account settings. Enable only the Apps needed for the task, review confirmations before consequential actions, and stop the task if its behavior looks suspicious. Prompt injection remains a risk, so avoid broad instructions such as “check my email and handle everything.”

See OpenAI’s Introducing ChatGPT agent and the ChatGPT agent Help Center article for current details.

Codex

Codex is the coding agent for writing, fixing, reviewing, testing, and shipping code. It goes beyond asking coding questions in chat: it can work with real repositories through the CLI, IDE, web, and app.

See Codex Entry Points for the detailed comparison.

Sora

Sora is OpenAI’s video generation and video workflow product. It may relate to the same account or credit model as ChatGPT, but its purpose is different from text chat and Codex.

Think of Sora for creating video material, exploring short video concepts, or drafting creative video assets.

Summary

• ChatGPT combines chat, search, files, data analysis, images, voice, Canvas, Projects, Tasks, GPTs, Apps, Deep Research, agent mode, and Codex
• Use search for current facts, Deep Research for longer investigations, and files/data analysis for documents and tables
• Use ChatGPT agent for multi-step online work, while limiting Apps, logins, and confirmation scope
• Use Projects for ongoing work, Canvas for editing, and Custom GPTs for repeatable use cases
• External tools and company information require Apps plus admin controls
• Real codebase work belongs in Codex workflows

See the references for the external specifications and background sources used on this page.[1][2]

References

1. OpenAI, Codex documentation
2. OpenAI, OpenAI API documentation