ChatGPT Projects vs Claude Projects: how they differ and which to use

What is the difference between ChatGPT Projects and Claude Projects?

Both are workspaces that group chats under shared instructions and files; the differences are in how they handle knowledge, memory, sharing and your local files. Claude Projects are built around a project knowledge base, with retrieval on paid plans when the files outgrow the context window. ChatGPT Projects keep chats, files, instructions and connected sources together, and the ChatGPT desktop app shows them next to local projects linked to folders on your computer.

Anthropic's help center says "Projects allow you to create self-contained workspaces with their own chat histories and knowledge bases." OpenAI's projects documentation says "Use a project to keep related chats, files, instructions, and sources together." Same idea, two implementations.

A note on sources: this comparison uses each company's official docs as read on 24 October 2026. OpenAI's projects page doesn't cover plans, file limits or sharing, so those rows point to OpenAI's help center instead of guessing.

How do Claude Projects work?

A Claude project has three parts: chats, project knowledge and project instructions.

• Knowledge: you upload documents, text, code or other files, which Claude uses as context for every chat in the project. Per Anthropic's file upload limits, project files can be up to 30MB each, and the number of files is "Unlimited, but total content must fit within Claude's context window."
• Retrieval on paid plans: when knowledge approaches the context limit, Claude "seamlessly enables RAG mode to expand capacity by up to 10x while maintaining response quality." This is only on paid plans (Pro, Max, Team and Enterprise).
• Instructions: project instructions tailor every chat in the project, such as a formal tone or a specific role. Per Anthropic's personalization article, they work alongside your account-wide instructions, which apply to all conversations.
• Plans: projects are available to all users, but "Free users can create a maximum of five projects."

Anthropic is also rolling out a new version of projects, starting with Claude Code, where a project is one conversation that Claude splits into parallel threads running in the cloud, each starting "with the project's files, repositories, instructions, and memory." The help center says it's in beta for select Pro and Max subscribers who use Claude Code, with chat, Cowork, Team and Enterprise to follow. For the full picture, see our Claude Projects guide.

How do ChatGPT Projects work?

A ChatGPT project groups related chats with shared context. OpenAI's guidance on when to make one: "Create a project when work will continue over time, produce more than one output, or depend on the same files and sources."

• Instructions: "Project instructions apply across its chats." OpenAI's personalization docs add that "Projects and repositories can also provide their own instructions," separate from your custom instructions.
• Sources: you add files and connected sources to the project's Sources section "when they should be available across its chats." Files attached to a single chat apply only to that chat.
• Plugins and MCP: the docs say you can install plugins to bring in context and actions from other services (on the web, in ChatGPT Work), and in the desktop app configure MCP servers where your organization or setup provides them.
• Local folders: on the web, "A ChatGPT project doesn't provide direct access to a folder on your computer." In the ChatGPT desktop app, the Projects view combines ChatGPT projects with local projects linked to folders, where Codex can work.
• Organizing: you can pin projects and chats, rename and archive chats, and search both.

For durable guidance in Codex projects, OpenAI's advice is notable: "Keep durable project guidance in AGENTS.md or checked-in documentation so it is available to future chats."

ChatGPT Projects vs Claude Projects side by side

How does memory work in each?

Claude keeps project memory separate. Anthropic's memory article says "Each project has its own memory." Inside a project, Claude uses that memory plus your memory from chats outside projects, but not memory from other projects. To keep a project fully separate, you turn off "Use account memory" on the project's page. Memory is on by default for Free, Pro and Max; on Team and Enterprise, owners decide whether it's available. Deleting a project deletes its memory.

ChatGPT describes memories more generally: they "let ChatGPT and Codex carry useful context from earlier work into future work," and you manage them under Settings, Personalization. OpenAI also says "Memories are separate from required project guidance."

Both companies land on the same advice: if something must always be true, write it into the instructions or a file. Don't rely on memory to pick it up.

How does team sharing compare?

On Claude, project sharing is available on Team and Enterprise plans. A project can be visible to everyone in the organization or shared with specific people as Can view (see contents and chat, but not edit) or Can edit (change instructions and knowledge, manage members). Your chats inside a shared project stay private unless you choose to share them.

ChatGPT also offers shared projects: OpenAI's ChatGPT Space docs note that files uploaded directly to a shared project "remain available to members." For which plans include sharing and any limits, OpenAI points to its Projects in ChatGPT help article, so check there for your plan.

When should you pick Claude Projects?

• Large reference material. A long spec, a codebase export or a stack of research papers fits better with retrieval on paid plans.
• Teams on Claude Team or Enterprise that want clear view and edit permissions per project.
• Strict separation. Per-project memory, with the option to cut off account memory, keeps client work apart.
• You already use Claude Code. The new projects beta brings repositories and parallel cloud threads into the same idea.

When should you pick ChatGPT Projects?

• Mixed sources. Projects can pull in connected sources and plugins alongside uploaded files, plus MCP servers in the desktop app.
• Desktop work with local folders. The desktop app puts ChatGPT projects and local Codex projects in one view.
• You already work in ChatGPT and Codex, and want project guidance to line up with an AGENTS.md in your repo.

Which should you pick?

Pick the app you already do the work in. The project features are close enough that switching for projects alone rarely pays off. If you use both, keep your project instructions and core prompts as plain text you own, so either app gets the same setup in a minute. Reusable procedures that span many projects fit better as skills: Anthropic's Skills explained post says "If you find yourself copying the same instructions across multiple Projects, that's a signal to create a skill instead." See Claude skills.

Where Kipt fits

Projects live inside each app. Kipt keeps the reusable layer outside them: your prompts, with {{name}} or {{name:default}} blanks, and your Agent Skills, as Markdown in a private GitHub repo you own, with tags and Cmd+K search. That makes it simple to paste the same instructions into a Claude project and a ChatGPT project, and to update both from one copy when you change them.