Claude Projects vs ChatGPT Projects: What Differs
Claude Projects and ChatGPT Projects do the same basic job: group related chats with shared files and instructions so you don’t re-explain your work every time. The difference is in how they carry context.
Claude leans on a project knowledge base that every chat reads, while ChatGPT leans on project memory that carries context between the chats inside a project. Neither shares anything with the other app, which matters if you use both.
What is a project in Claude and in ChatGPT?
In both apps, a project is a workspace: a set of chats plus files and instructions that apply to every chat inside it.
Claude’s help center describes projects as a way to group chats around shared knowledge. Anything you upload to the project knowledge base “will be used across all of your chats within that project,” and project instructions apply to every chat.
OpenAI’s help center describes projects in ChatGPT as workspaces that keep everything for a long-running effort in one place. Projects also have built-in memory of the chats and files inside them.
How do Claude Projects and ChatGPT Projects compare?
This table reflects each company’s help pages at the time of writing. Plans and limits change often, so check the linked pages before you decide.
| Claude Projects | ChatGPT Projects | |
|---|---|---|
| Who gets them | All users; free accounts can create up to five | All free and paid plans |
| Instructions | Project instructions apply to every chat in the project | Project instructions apply inside the project and override your global custom instructions |
| Files | Project knowledge base used by every chat | Uploaded files, plus links from supported apps such as Google Drive and Slack |
| Large file sets | Paid plans switch to retrieval (RAG) mode as knowledge nears the context limit | Depends on plan file limits |
| Memory between chats | Chats don’t share context unless it’s in the knowledge base or instructions; projects keep their own memory, separate from non-project chats | Project memory remembers the chats and files in the project; an option keeps memory scoped to the project only |
| Sharing | Team and Enterprise plans | Business, Enterprise and Edu plans |
| Moving chats in | Yes | Yes, except chats made with a custom GPT |
How does memory work inside each kind of project?
This is the biggest practical difference, and the one people get wrong.
Claude is explicit that chats in a project don’t see each other by default. Its help page says context “is not shared across chats within a project unless the information is added into the project knowledge base.” The same page says each project has its own memory, kept apart from chats outside projects. So the reliable way to give every Claude chat a fact is still to put it in the knowledge base or the instructions.
ChatGPT projects remember the chats and files inside them, so a new chat can pick up on what you discussed in an earlier one. Coverage such as TechRadar’s also describes a project-only memory setting that keeps a project from drawing on your wider ChatGPT memory, and keeps what it learns from leaking out. Reports differ on whether you can switch it on for an existing project, so look in the project’s settings.
If you want the account-level version of this comparison, Claude memory vs ChatGPT memory covers how each app remembers you outside projects.
Which is better for writing and content work?
For long-form writing with a fixed brief, Claude Projects work well. Put the brand guide and two sample pieces in the knowledge base and set instructions for voice. Every chat then starts from the same material.
For ongoing work that changes week to week, ChatGPT Projects have the edge, since project memory picks up decisions from earlier chats without you adding them to a file. The catch is that you can’t easily see or edit what the project memory holds.
Say you write newsletters for two clients. In either app you’d make one project per client. With Claude, you’d update each client’s knowledge base when something changes. In ChatGPT, you’d rely more on the project remembering, and check in now and then that it remembered the right things.
For the brand-voice side of this, see how to make ChatGPT remember your brand voice.
Which is better for coding and research?
For research, both work. Claude’s retrieval mode on paid plans helps when you load many long documents. ChatGPT’s app links help when your sources live in Google Drive.
For coding, many developers work in an editor such as Cursor or VS Code, or in a terminal agent such as Claude Code, rather than in a chat project. Those tools keep their own context, and none of it is the same as your chat project.
What can’t either kind of project do?
Three gaps apply to both:
- They don’t cross apps. A Claude project is invisible to ChatGPT, and the other way around. Your editor sees neither. If you switch between AIs, you maintain the same context twice. Sharing memory between Claude and ChatGPT explains the workarounds.
- Files are snapshots. You upload a brief in March. In May the brief changed, and the project still has the March version until you replace it by hand.
- Knowledge doesn’t grow on its own in a form you can read. Claude’s knowledge base only holds what you upload. ChatGPT’s project memory grows, but it isn’t a set of documents you can open and edit.
FAQ
Are Claude Projects free?
With limits. Claude’s help page says free users can create up to five projects. Sharing projects requires a Team or Enterprise plan.
Are ChatGPT Projects free?
OpenAI lists projects as available on all free and paid plans. Sharing with teammates needs a Business, Enterprise or Edu workspace.
Can I move a project from ChatGPT to Claude?
Not directly. Copy your instructions into a new Claude project and upload the same source files from your own copies. Chat history and project memory don’t transfer.
Do project instructions replace custom instructions?
In ChatGPT, project instructions override your global custom instructions inside that project. In Claude, project instructions apply to every chat in that project, so put project-specific rules there.
Where Valta fits
If you use only one AI, its projects are probably enough. If you use Claude for drafting, ChatGPT for research and Cursor for code, Valta gives them one shared place to work from. It’s a hosted vault of Markdown notes that each AI connects to over MCP. Any of them can save a decision, and any of them can find it later with its linked notes. On Pro you can keep up to five vaults, one per client or brand (see managing multiple brands with AI), each with a profile your AIs read first, much like project instructions that every AI shares.
Valta doesn’t replace project files for one-off uploads, and it runs no AI of its own. It keeps the notes that should outlive any single chat. Start with a free vault.