Memory and context

ChatGPT Projects vs Claude Projects: which works better for a team?

ChatGPT Projects and Claude Projects compared for teams: who sees which chats, file limits, how memory works in shared projects, and what each plan costs.

Choose ChatGPT Projects when teammates need to inspect and branch project chats. Choose Claude Projects when teammates need shared files and instructions with private conversations by default. In a shared ChatGPT project, members can see and branch each other’s chats; in a shared Claude Project, your chats stay private unless you share them. ChatGPT Business limits each project to 40 files, while Claude can switch to retrieval when knowledge grows. A shared ChatGPT project is permanently set to project-only memory, while a Claude Project has its own memory that also uses your personal account memory unless you turn that off. Standard seats on both plans cost $25 per user per month when billed monthly, or $20 per month when billed annually, excluding applicable taxes.

By Bombos. Published November 17, 2026. Sources checked November 16 and 17, 2026.

How do the two compare?

Aspect ChatGPT Projects (Business) Claude Projects (Team)
Standard seat price per user, as of November 2026, excl. taxes $25 monthly, $20 billed annually $25 monthly, $20 billed annually
Team size From 2 users 2 to 150 seats
Sharing available on Every plan, including Free and Plus Team and Enterprise only
Share with People, groups, or the whole workspace by link Invited members, or everyone in the organization
Roles Owner, Edit, Chat Can view (includes chatting), Can edit
Colleagues’ chats Visible to all members Private unless shared
Files per project 40, up to 10 per upload No fixed number; RAG up to 10x capacity
App sources Google Drive and Slack channel links, in a private project Connectors included in the Team plan
Memory in a shared project Project-only, cannot be switched back Own project memory; account memory can be disabled separately
Collaborators per project Up to 100 Not stated in the help articles

Sources: OpenAI Help Center, Projects in ChatGPT, OpenAI Help Center, ChatGPT Business overview, Claude Help Center, What are projects?, Claude Help Center, Project visibility and sharing, claude.com/pricing, read November 16 and 17, 2026.

Who sees which chats?

Start with the question of who may see each other’s conversations.

In ChatGPT, members can inspect and branch project chats. OpenAI says members can “pick up where others left off”; they cannot collaborate synchronously in one chat. You can see who started a chat by its profile icon, and members can branch a teammate’s chat “to explore a new idea without losing the original thread”. Members can archive or delete chats they own, after which others no longer see them (OpenAI Help Center). That suits a deal, a client file or a report three people work on together.

Claude is a shared library. Sharing a project “will share the project and knowledge base with the member, but your chats will remain private unless you choose to share them”. This holds even when the project is open to everyone in the organization (Claude Help Center). Everyone works from the same files and instructions, each in their own conversations.

How much knowledge fits?

ChatGPT is explicit: “Edu, Pro, Business, Enterprise: 40 files per project”, and “only 10 files can be uploaded at the same time”. Projects are unlimited. Private projects can also add Google Drive files and folders and Slack channels as sources, but “The Google Drive app does not support sync when added within a project” (OpenAI Help Center).

Claude gives no count. When project knowledge approaches the context limit on a paid plan, Claude “seamlessly enables RAG mode to expand capacity by up to 10x”. Free users can create at most five projects (Claude Help Center).

In practice: Claude can enable retrieval as project knowledge approaches the context limit, and then it reads what retrieval finds, so a question that needs two facts from different files can miss one. ChatGPT Business limits each project to 40 uploaded files; file count alone does not establish which product holds more usable information. Either way, an uploaded file is a copy that goes stale when the original changes.

How does memory work in a shared project?

ChatGPT has default memory and project-only memory. “Shared projects are automatically set to project-only memory and cannot be switched to default memory”, and they “do not have access to an individual member’s context, custom instructions, or memories outside the project”. For Business, this happens “at the time of sharing, regardless of any previous memory setting”. ChatGPT Work is not available in such a project (OpenAI Help Center).

Claude gives each project its own memory: “Each project has its own memory.” Inside a project, Claude also uses memory from your chats outside projects, never from other projects, unless you turn off “Use account memory” for that project. On Team, owners decide whether memory is available, and “Owners can’t view or edit a user’s individual memories” (Claude Help Center). Anthropic does not describe whether project memory is shared between members of a shared project.

Personal memory is not automatically shared team knowledge. Within a shared ChatGPT project, however, project conversations can influence later answers. Rules for everyone belong in the project instructions. See How do you make AI agents learn from corrections?.

What do admins control?

In ChatGPT Business, workspace owners decide who may create and who may share projects under Workspace settings, Permissions & roles. Projects inherit workspace settings: if memory or deep research is off for the workspace, it is off in every project, and messages follow the workspace retention period. “By default, OpenAI does not use content from ChatGPT Business, Enterprise, Edu, or ChatGPT for Healthcare workspaces to train its models” (OpenAI Help Center).

In Claude Team, owners turn memory on or off for the organization under Organization settings, Capabilities. A project is either private to invited members or public to the whole organization, switched from the Share button (Claude Help Center).

Which fits your team?

  1. Several people work in the same case and want to see each other’s conversations: ChatGPT Projects.
  2. Everyone should use the same style guide, price list and process, each in their own chats: Claude Projects.
  3. Lots of documents per topic: Claude, or merge files into fewer, larger ones in ChatGPT.
  4. Developers already use Claude Code or Codex: that choice often decides the rest. See ChatGPT Business vs Claude Team.

If your team uses both, knowledge and instructions live in two places. Project content is not automatically synchronized between ChatGPT and Claude. For the coding side, see Claude Code memory for teams. If you are looking beyond projects to search across company apps, see Glean alternative for startups and small teams.

Read also: keeping company context portable across ChatGPT, Claude and Copilot.

How does Bombos do this?

A project is good shared context for one topic, inside one tool. Bombos is the memory layer under the AI a team already uses: Claude, Claude Code, ChatGPT, Copilot, Codex, Cursor or your own agents. It is a plugin, not a new chat or model.

  • One shared memory that learns how the team works, available to the connected tools within the permissions you set, rather than per project or per subscription.
  • The source behind every answer.
  • Your AI acts, Bombos remembers. Bombos does not take actions itself.
  • Cloud on servers in Europe, or standalone on your own server. One person can try the plugin for their own agents first.

Where Bombos is not the answer: if your whole team works in one tool and one project per client covers what you need, the built-in project is enough.

Frequently asked questions

Can I share a Claude Project on Pro? No. Project sharing is available on Team and Enterprise, with members of your organization.

Do colleagues see my chats in a shared ChatGPT project? Yes. Members can see and work with the chats, files and instructions in the project. Your chats outside the project stay private.

Can a shared ChatGPT project go back to default memory? No. Even if all collaborators are removed, it stays on project-only memory.

How many files fit in a Claude Project? Anthropic gives no fixed number. On paid plans Claude switches to retrieval when knowledge nears the context limit, for up to ten times more capacity.

Book your half hour: together we map where your team’s knowledge lives now, and what your AI misses. Free, and the plan is yours, with or without Bombos.

Which sources did we use?

All sources opened and read on November 16 or 17, 2026.

  1. OpenAI Help Center, Projects in ChatGPT: https://help.openai.com/en/articles/10169521-projects-in-chatgpt
  2. OpenAI Help Center, ChatGPT Business overview: https://help.openai.com/en/articles/8792828-chatgpt-business-overview
  3. Claude Help Center, What are projects?: https://support.claude.com/en/articles/9517075-what-are-projects
  4. Claude Help Center, Project visibility and sharing: https://support.claude.com/en/articles/9519189-project-visibility-and-sharing
  5. Claude Help Center, Use Claude’s chat search and memory to build on previous context: https://support.claude.com/en/articles/11817273-use-claude-s-chat-search-and-memory-to-build-on-previous-context
  6. Claude pricing: https://claude.com/pricing
The first half hour · free

What does your AI forget?

  1. 01You tell us which AI you use and where it keeps starting over.
  2. 02We show what a shared memory changes there, on your own examples.
  3. 03You keep the plan, with or without Bombos.
Piet Baudoin
Piet BaudoinYou talk to Piet, from the Bombos teamDutch team from Utrecht and Nijmegen · data in Europe
We reply the same working day.