How to Use ChatGPT Projects to Organize Your Work (A Practical Guide)

What ChatGPT Projects actually are

If you use ChatGPT for real work, you have probably felt the sidebar problem. You did something useful three weeks ago, you know it is in there somewhere, and you cannot find it. Meanwhile every new chat starts from zero, so you paste the same background in again and again.

Projects fix that. OpenAI describes them as smart workspaces that keep everything tied to a long-running effort in one place: you group chats together, upload reference files, and add custom instructions, so ChatGPT remembers what matters and stays on topic. [1] Think of a Project as a folder with a brain. Everything you put inside it shares context, so the assistant stops forgetting where you left off.

They are available on every subscription type, free and paid, and they do not cost anything extra beyond your normal plan. [1] That combination, useful and free, is why I now tell almost everyone I train to set one up before they do anything clever with prompts.

How to set up your first project in two minutes

There is no trick to this. The whole point is that it is fast.

1. Click New project in the ChatGPT sidebar. [1]
2. Give it a name and pick an icon and colour so you can spot it quickly.
3. Open the project and start a chat, or drag an existing chat into it.
4. Add your reference files and a line or two of project instructions (more on both below).

That is genuinely it. From now on, any chat you start inside that project inherits its files and instructions automatically. You will notice the difference on the second conversation, when ChatGPT already knows the context you would normally have to re-explain.

If you have older chats that belong in the project, you do not have to recreate them. Drag a chat onto the project, or open its menu and choose Move to project, and it inherits the project’s instructions and file context. [1] One thing to know: chats you started with a Custom GPT cannot be moved in, so start those fresh inside the project instead. [1]

Project instructions: the setting most people skip

This is the feature that separates people who like Projects from people who love them, and almost everyone misses it at first.

Click the three dots in the top right of your project and choose Project settings to add project instructions. [1] These are standing orders that apply to every chat inside that project, and crucially, they override your global custom instructions. [1] So you can have a research project that always replies with sources, and a writing project that always matches your tone, without changing your account-wide settings.

Keep them concrete. OpenAI’s own example is the right length: “Act like my marketing mentor. Be concise. Use bullet points. Ask clarifying questions.” [1] If you have never set custom instructions at all, start with our ChatGPT custom instructions setup guide, then apply the same thinking at the project level.

Adding files (and the limits that bite)

You can add PDFs, spreadsheets, docs, images, or pasted text as reference material, and ChatGPT will draw on them in its answers. [1] This is what makes a project feel like it actually knows your work rather than guessing at it.

The number of files you can add depends on your plan, and the limits are worth knowing before you hit one mid-task: [1]

• Free: 5 files per project
• Go and Plus: 25 files per project
• Pro, Business, Enterprise, and Edu: 40 files per project

You can only upload 10 files at a time, so larger sets go up in batches. [1] If you are on Business or Enterprise, you can also paste links from Google Drive files and folders or Slack channels as project sources, which keeps the project working from the same up-to-date material your team already uses. [1]

Hit the limit? The fix is housekeeping, not a bigger plan: remove old uploads, combine files, or split the work across two projects. [1]

Project memory, explained without the jargon

Projects have built-in memory, which simply means the project remembers the chats and files inside it so you can pick up where you stopped. [1] There is one choice worth understanding, because it trips people up.

When you create a project, you pick whether its memory is project-only or default. [1]

• Project-only memory keeps the project sealed off. Chats inside it can reference each other but cannot see your other ChatGPT history, and your saved memories from outside are not pulled in. [1] This is ideal for sensitive or focused work where you want the assistant anchored to that project alone.
• Default memory lets the project also draw on your wider saved memories and, on consumer plans, other conversations. [1]

The catch: you choose this when you create the project, and an existing project on default memory cannot be switched to project-only later. If you want a sealed workspace, start a new project for it. [1] You can still move conversations between projects if you need to reshuffle.

Sharing a project with your team

Since October 2025, project sharing is available to all ChatGPT users, including Free, Plus, Pro, and Go, on web, iOS, and Android. [1] A shared project becomes a live context hub: ChatGPT can draw on the project’s chats, files, and instructions so responses reflect your group’s shared knowledge, and anyone can pick up where a colleague left off. [1]

Use the Share button in the top right to invite people. Invitees get one of two access levels: [1]

• Edit access lets members update instructions, add or remove files, and invite others.
• Chat access lets members see and use the project’s chats, files, and instructions, but not invite others.

A few things worth flagging for work use. When a project is shared, its memory automatically becomes project-only, so it cannot reach any member’s private context outside the project. [1] And the data rule is sensible: OpenAI says it will only train on shared-project data if every contributor and the owner have the “Improve the model for everyone” setting turned on. [1] For Business, Enterprise, and Edu accounts, this is not used for training by default. [1]

Project or Custom GPT? How to choose

People conflate these constantly, so here is the clean distinction OpenAI draws. [1]

A Custom GPT is a reusable, single-player tool: a curated expert you build once and call again and again, like an “HR Benefits GPT” that always answers policy questions the same way. A Project is a multi-player, evolving workspace: a place where context builds up over time as you and your team add chats, files, and instructions, like a “Q4 Planning” hub. [1]

The rule of thumb: if the thing you want is stable knowledge you will reuse across many situations, build a Custom GPT. If the thing you want is a living home for one ongoing effort, use a Project. And if you are deciding between ecosystems entirely, our guide to how Claude Projects works covers the closest equivalent on the Anthropic side.

Five projects worth setting up today

Abstract features are easy to forget, so here are concrete starting points that pay off within a week:

• Weekly reporting. Store your data and last week’s report; return each week to generate the update without re-explaining anything. [1]
• A recurring client or account. Keep contracts, notes, and brand details in one place so every draft is on-brand.
• A research topic. Keep transcripts, survey results, and sources together and query across them. [1]
• A writing project. Set project instructions for your voice and let every draft start in the right tone.
• Onboarding or a study guide. Upload the key documents and use it as a patient explainer you can ask anything.

Each of these works better once you can also write a sharp prompt, so if your results feel flat, run them through the 4-part prompt formula. And if you want your whole team fluent in this kind of practical setup, that is exactly what our AI courses for non-technical professionals are built to do.

Sana Mian, Co-Founder of Future Factors AI

Sana is an AI educator and learning designer specialising in making complex ideas stick for non-technical professionals. She has trained 2,000+ learners across corporate teams, bootcamps, and keynote stages. Future Factors offers AI Bootcamps, Corporate Workshops, and Speaking & Consulting for businesses ready to adopt AI without the overwhelm.

More about Sana →