Claude Projects vs ChatGPT Projects

Claude Projects and ChatGPT Projects solve the same problem: they keep the files, standing instructions and conversations for one piece of work in one place, so you stop re-explaining it. The structural differences are in sharing and memory. Claude shares projects only on its Team and Enterprise plans, and members see the knowledge and instructions but not each other’s chats. ChatGPT offers project sharing on its individual plans too, and members of a shared project see and continue each other’s chats. Knowledge, instructions and tools are closer than the marketing suggests.

At a glance

• Knowledge: both take uploaded files. Claude calls it project knowledge; ChatGPT calls them project sources and also accepts links to Google Drive files and Slack channels.
• Instructions: both have project instructions that apply to every chat in the project.
• Memory: Claude gives each project its own memory, separate from other chats. ChatGPT lets you choose default memory or project-only memory per project.
• Sharing: Claude on Team and Enterprise, with Can view and Can edit. ChatGPT on individual and business plans, with chat or edit access and an optional link.
• What collaborators see: in Claude, the knowledge and instructions; chats stay private unless shared one by one. In ChatGPT, the project’s chats as well.
• Tools: both let you use connected apps or connectors inside a project, with conditions described below.

Knowledge and files

In Claude, you upload documents, text or code to a project’s knowledge, and Claude uses it in every chat in that project. Anthropic’s help page What are projects?↗ (opens in a new tab) says that when project knowledge approaches the context limit, paid plans switch to retrieval, which expands capacity by up to ten times.

In ChatGPT, OpenAI’s help centre describes project sources: uploaded PDFs, spreadsheets, documents, images and pasted text, with the number of files depending on your plan. A private project can also take a link to a Google Drive file or folder or a Slack channel as a source, and you can save a useful chat response back into the project’s sources.

Instructions

Both tools give a project its own instructions: tone, format, the role to take, what to avoid. In ChatGPT, project instructions apply only inside that project and override your global custom instructions. In Claude, project instructions sit alongside the project knowledge and shape every chat in the project. The practical advice is the same for both and is covered in Claude Projects best practices: keep instructions short, write rules rather than background, and put reference material in the files.

Memory

This is where the designs diverge most. According to Anthropic’s article on chat search and memory↗ (opens in a new tab), each Claude project has its own separate memory space and project summary, kept apart from other projects and from chats outside projects. Moving a chat into or out of a project changes which memory it feeds. Without memory, Claude’s help centre notes that context is not shared across chats in a project unless it is added to the project knowledge.

ChatGPT gives each project a setting. With default memory, your saved memories apply and, on plans other than Enterprise and Edu, chats can draw on conversations inside and outside the project. With project-only memory, saved memories are not used, chats reference only conversations in the same project, and nothing outside the project can see in. Shared projects are always project-only, and a project cannot be switched back once shared.

Sharing and collaboration

Claude’s project sharing is for organisations. On Team and Enterprise plans, a project can be public to the organisation or private to invited members, and each member gets Can view or Can edit, as set out in Manage project visibility and sharing↗ (opens in a new tab). Sharing a project shares its knowledge and instructions; your chats in it stay private unless you share a chat. The full detail is in can Claude Projects be shared?

ChatGPT’s project sharing reaches further. OpenAI’s help centre lists it on Free, Go, Plus and Pro as well as Business, Enterprise and Edu. Invitees get chat access, which lets them see and use the project’s chats, files and instructions, or edit access, which also lets them change instructions and files and invite others. An owner can make the project open to anyone signed in to ChatGPT who has the link. Members can branch a colleague’s chat and continue it, and see who started each one.

The consequence is a different habit. In a shared Claude project, the shared thing is the reference material; each person’s thinking stays in their own chats. In a shared ChatGPT project, the chats themselves are the shared record, which is closer to a team room, and means anything typed in a project chat is visible to its members.

Connectors and tools

Both let a project reach outside itself. On Claude’s Team and Enterprise plans, Anthropic’s page on connectors↗ (opens in a new tab) says connectors are available only in private projects, each person authenticates separately, and chats with synced content cannot be shared. In ChatGPT, you can use connected apps in project chats from the tools menu, and ChatGPT may ask you to confirm before an app searches outside the project. Image generation, voice and web search work in projects as they do outside them, with other tools depending on the plan.

Both also accept custom connectors built on the Model Context Protocol, where your plan allows them, which is how either can be pointed at a tool of your own. OpenAI’s guide to projects↗ (opens in a new tab) frames them as dedicated spaces that gather context, files and instructions; the connectors are what let that space act on something outside it.

Moving existing chats in

Both let you gather conversations you have already had. In Claude you open a chat’s menu and choose “Add to project”, and the same menu removes it or moves it to another project; with memory on, that also decides which memory the chat feeds. In ChatGPT you drag a chat onto the project or choose “Move to project”, and the chat inherits the project’s instructions and files. Chats created with a GPT cannot be moved into a ChatGPT project, so for those you start a new chat inside it.

Which fits which job

• A body of reference material you and a colleague both query, where each of you wants your own working space: a Claude project on a Team plan.
• A small group, not on a business plan, who want to work in the same conversations: a shared ChatGPT project.
• Keeping a client’s work entirely apart from everything else you do: either, with ChatGPT’s project-only memory or Claude’s per-project memory.
• Work that lives in a codebase rather than in documents: neither; see Claude Projects vs Claude Code.

What neither one is

Neither project type is a list of work with owners. A project remembers what you told it; it does not record who is doing which task, what state it is in, or who changed it. When the conversations start producing “we should fix X” and “can you check Y”, that list belongs somewhere both people and assistants can read and update. fenbs is built for that: a board where each task is a feature, enhancement or bug in lanes To Do, Next Up, In Progress and Completed, and where Claude and ChatGPT can both connect over MCP, each holding a role, with every change recorded under who made it.