What Is the ChatGPT Projects Equivalent in Gemini?

Gemini does not have a feature called Projects. The closest equivalents are Gems (custom Gemini configurations, renamed Skills on personal accounts from November 2026) for ongoing instructions and personas, and Gemini Notebook (formerly NotebookLM) for deep research with source documents. Neither matches ChatGPT Projects exactly, and understanding the difference helps you choose the right approach for your work.

If you use both ecosystems or plan to switch between them, tools like fenbs keep the underlying work — tasks, decisions and progress notes — outside any one AI app, so the project context travels with you rather than staying locked in either platform.

What ChatGPT Projects do

ChatGPT Projects let you group conversations under a shared name, carry custom instructions and uploaded files across every conversation in the project, and pick up where a previous session left off within the same project. The project holds the context; individual conversations inside it all start from the same shared instructions and knowledge.

For an ongoing engagement — a client account, a codebase, a research area — Projects reduce the setup at the start of each session. You open the project instead of pasting a long context again. For details on shared ChatGPT projects, see chatgpt shared projects.

Gems and Skills: Gemini's persistent configurations

Gems are custom Gemini setups: a name, a system prompt and any background context you want applied whenever that Gem is the active assistant. You can create multiple Gems for different roles — one for legal drafting, one for code review, one for customer emails. From November 2026, Gems are renamed Skills on personal Gemini accounts; the underlying function is the same.

Gems carry instructions across conversations, which is similar to the custom instructions in ChatGPT Projects. They do not group conversations together or show a history of what was worked on across sessions. Each conversation with a Gem starts from the instructions you set, not from where the last conversation ended.

Google Workspace users with Gemini for Workspace access may have additional ways to organize work across conversations, depending on their plan. Check your Workspace admin or billing page↗ (opens in a new tab) for what is available to your account.

Gemini Notebook: deep research with sources

Gemini Notebook (formerly NotebookLM, renamed July 2026) is a research tool built around source documents. You upload PDFs, web links, Google Docs and other sources, and Gemini Notebook answers questions grounded in those specific materials. It cites where in your sources each answer comes from.

Gemini Notebook is well suited to synthesizing a body of research — a set of case studies, a collection of reports, the documents for a project proposal. It is less suited to managing ongoing task work or carrying active instructions across sessions. Think of it as a library tool rather than a workspace tool.

For more on Gemini Notebook and what it can do, see the Gemini Notebook overview.

Memory across conversations

Gemini has a memory feature on some plans that lets it remember details you share across conversations — preferences, recurring facts and context it has picked up over time. This is different from Projects: memory is about personal context that follows you everywhere, while Projects are about organizing a specific body of work.

ChatGPT also has memory, operating in a similar way. For a comparison of the two memory systems, see Gemini memory vs ChatGPT memory. Neither system replaces a deliberate project record: an AI's memory reflects what it noticed, not what you chose to record.

When Gems or Skills fit best

Use a Gem or Skill when you want the same role or set of instructions applied every time you open a conversation in Gemini — a tone, a format, a persona, a domain focus. The value is in the setup you write once and reuse, not in organizing work across sessions.

Gems and Skills work well for recurring task types rather than specific projects. A legal drafting Gem that always follows a particular style is a good use of the feature. A Gem tracking the status of one client project is not, because it cannot update itself as the project changes.

When Gemini Notebook fits best

Use Gemini Notebook when your work is to understand, summarize or question a defined set of documents. Legal document review, academic research, proposal preparation from a stack of client materials — these are strong fits. You bring the sources; Gemini Notebook helps you get answers from them.

It is a weaker fit for managing the task side of work: what to do next, who agreed to what, what rules apply. Those are better held in a task record you maintain and update rather than in a document-question interface.

Migrating context from ChatGPT Projects to Gemini

If you are moving ongoing work from ChatGPT Projects to Gemini, the process is to extract the key facts explicitly: the project goal, the custom instructions, the decisions made so far and where work stands. None of that moves automatically — you export or copy it and set it up in the Gemini equivalent.

For instructions that should persist across all future Gemini conversations on this work, put them in a Gem or Skill. For the task state and decisions, put them in fenbs or another record outside either platform. That way, if you move again or use both platforms in parallel, the core context is not locked in either product.

Cross-platform workflows in practice

In practice, many people use both ChatGPT and Gemini for different tasks on the same project. ChatGPT Projects may hold the active task context while Gemini Notebook handles document research for the same work. This is fine as long as there is one authoritative record of what was decided and where things stand.

The friction comes when the authoritative record is split: some decisions live in ChatGPT's project instructions, others in a Gemini Gem's system prompt, and neither has the full picture. When you start a new session in either product, you begin from a partial view. Moving decisions and task state to a shared record outside both tools avoids this split.

One practical approach: keep ChatGPT Projects or Gemini Gems for role-specific instructions that do not change often — tone, format, professional style — and keep task state, decisions and where work stands in a record both tools can read. The instructions in the AI product describe what it is; the shared record describes what happened.

Also worth noting: custom Gemini Gems are not a permanent name. From November 2026 they become Skills on personal accounts. If you are writing documentation about your setup, use "Skills" for anything you expect to last beyond that date. The function — custom persistent instructions — is the same, and existing Gems carry over without any action needed from you.

A useful rule of thumb: if the instruction answers "how should the assistant behave," put it in the AI product's configuration. If the instruction answers "what did the project decide," put it in the shared record. Those two categories rarely need to be in the same place.

How to do this with fenbs

1. Create a free account and add the project you are organizing.
2. Write the project goal, the tasks that make it up, and any decisions or rules that apply across sessions.
3. Connect your preferred AI app — Gemini, Claude or ChatGPT — over MCP from Integrations.
4. At the start of each session, ask the assistant to read the project context from fenbs. It gets the task state, rules and where work was left, independent of which app you are using.
5. Before stopping, ask the assistant to save Where we left off so the next session on any platform starts from the current state.

For a broader comparison of ChatGPT and Gemini across tasks, see ChatGPT vs Claude vs Gemini. For advice on best practices inside ChatGPT Projects, see chatgpt projects best practices.