Best ChatGPT Features You're Probably Not Using

You're Probably Using About a Fifth of What's There

ChatGPT has grown from a text box into something closer to a small operating system — memory, project folders, study tools, custom behavior settings, connected apps. Most of that growth happened in small, unannounced updates buried in release notes nobody reads. The result is a huge gap between what ChatGPT can do and what an average user actually asks it to do.

None of the features below require a hidden trick or an unofficial workaround. They're all sitting in the standard interface, most of them behind Settings → Personalization or the tools menu next to the message box. The only reason they go unused is that nothing points you to them — OpenAI tends to ship these quietly, mention them once in a changelog, and move on, which means even people who've used ChatGPT daily for a year can still be missing half of what's already available to them for free.

Custom Instructions Set the Rules; Memory Learns the Rest

Memory carries context between conversations automatically, without you re-explaining who you are every time.

Memory lets ChatGPT remember details across separate conversations — your job, your ongoing projects, your writing preferences, even your name — instead of starting from zero every time you open a new chat. It builds up gradually as you use the product, but you can also manage it directly: go to Settings → Personalization → Memory to see, edit, or delete individual memories, or wipe everything and start clean.

Two settings worth knowing sit right next to it. If you want a single conversation to not affect your stored memory at all, use Temporary Chat — toggle it on from the model selector for a session that leaves no trace and pulls from no saved context. If you just need to override something for one message without deleting it permanently, start that message with "For this conversation only..." and ChatGPT will treat the override as scoped to that exchange.

Stop Typing the Same Preamble Every Time

If you find yourself typing the same context or formatting request at the start of every chat, that belongs in Custom Instructions instead. Go to Settings → Personalization → Custom Instructions, and you'll see two boxes: one for who you are and what you're working on, the other for how you want responses formatted.

A useful pattern: in the first box, describe your role and what you're generally trying to accomplish ("I run a small e-commerce business and handle marketing, customer emails, and product descriptions"). In the second, specify format preferences ("Default to bullet points for steps, keep tone professional but plain, always include a one-line summary at the top"). Once saved, every new conversation follows those rules automatically — no re-explaining, no repeated setup.

Keep Related Chats and Files Together

Projects group related conversations, uploaded files, and custom instructions into one dedicated space — useful the moment you're juggling more than one ongoing piece of work. A freelancer might keep a Project per client; a student might keep one per course. Every chat inside a Project can reference the files you've uploaded there without you re-uploading them each time, and project-level instructions apply automatically to every conversation inside it.

This matters more than it sounds, because it solves the single most common complaint about long-running AI use: losing track of which chat had which context. Instead of scrolling through dozens of unrelated conversations in your sidebar, related work stays grouped and searchable.

An Actual Tutor Mode, Not Just an Answer Machine

Study Mode changes how ChatGPT responds to learning questions: instead of handing you a finished answer, it asks what you already know, walks you through the problem in steps, and checks your understanding along the way — a Socratic approach rather than a lookup. Turn it on from the + tools menu in the message box (or search "Study" if it's not visible), and it applies for that conversation. It's not available inside Temporary Chats, GPTs, or Projects, so if it's missing, that's usually why.

It's genuinely different from just asking ChatGPT to "explain this" — Study Mode is built specifically to slow the interaction down and force some of the thinking to stay with you, which is the opposite of what most people reach for AI to do, and exactly why it's worth trying if retention actually matters to you.

Let It Actually Browse, Not Just Answer From Memory

Deep Research is an autonomous research mode that plans out sub-questions, browses dozens of sources, reads full articles and PDFs, and returns a structured, cited report — work that would otherwise take hours of manual searching. It's available from the tools menu on ChatGPT Plus and higher, with usage limits that vary by plan (Free users get a small monthly allowance; Plus, Team, and Enterprise get more; Pro gets the most). You can also restrict Deep Research to specific trusted sites, or connect it to your own files through Google Drive, OneDrive, Dropbox, or SharePoint, so it researches against your own documents instead of only the open web.

Two Smaller Habits Worth Building

Archived Chats is easy to miss but worth using regularly: any conversation you archive disappears from your sidebar without being deleted, so you can declutter without losing history you might need later (Settings → Archived chats brings it all back). And under Settings → Connected apps, you can link Google Drive, Microsoft OneDrive, and similar services so ChatGPT can pull in your own documents directly instead of you copy-pasting content into the chat window every time.

None of these five main features or two smaller habits require a technical background to use — they're all sitting in menus most people never open. The gap between "using ChatGPT" and "getting real value from ChatGPT" is usually just a few minutes spent in Settings once.

Where to Turn These On

None of these require an upgrade to try. Start with Custom Instructions — it takes under two minutes and changes every conversation from that point forward — then layer in Memory and Projects once you notice which conversations you keep re-explaining yourself in.

Frequently Asked Questions