Google Tests Projects for Gemini Desktop: Persistent Workspaces Coming

Google is internally testing a Projects feature for the Gemini desktop app. The interface creates dedicated workspaces where users set custom instructions, attach reference files, and keep related chats together in one place. The feature is not publicly available, no release date has been confirmed, and details like account eligibility and cross-device sync remain unknown. It mirrors what ChatGPT and Claude have offered for a while, and what Gemini Business already supports, but would bring that structure to consumer desktop users for the first time.

What happened

Detail	Status
Feature name	Projects
Platform	Gemini desktop app
Current availability	Internal testing only
Public release date	Not confirmed
Account eligibility	Unknown
Cross-device support	Unknown
Existing availability	Gemini Business (longer-standing)

Google is running internal tests on a functional Projects interface for the Gemini desktop app. Each project acts as a self-contained workspace: users write custom instructions, upload reference files, and every chat started inside that project inherits that shared context automatically.

The feature was identified through a working interface in internal testing. Google has not made a public announcement, so a consumer rollout is not guaranteed, though the working state of the interface suggests broader availability is planned.

How Projects differ from Gemini Notebooks

This is an important distinction worth noting. Gemini already has a Notebooks feature tied to NotebookLM, which also lets users collect sources, add instructions, and hold ongoing conversations. Projects are separate and follow the structure that ChatGPT and Claude users will recognise: one persistent workspace where all new chats share the same base context.

Google’s public Enterprise documentation already describes Projects with a dedicated knowledge boundary, uploaded and linked files, and multiple private chats per project. The consumer Gemini documentation, by contrast, currently points users toward Notebooks for similar needs, with Mac app support listed.

Why does this matter for everyday users?

Right now, Gemini consumer users have no straightforward way to keep a project’s context alive across multiple chat sessions without manually re-uploading files or re-entering instructions each time. Projects would fix that in one step.

The use cases are obvious: a developer maintaining a coding assistant with project-specific rules, a writer keeping a style guide attached to every draft session, a researcher grouping all related conversations in one place. Teams managing recurring work across several sessions would benefit most.

If you already use AI integration in your business workflows, persistent project context is one of the most practical quality-of-life improvements an assistant can offer. It reduces setup friction and makes the tool more reliable across long-running tasks.

Our take

This is a catch-up move, and there is nothing wrong with that. ChatGPT introduced Projects in late 2024 and Claude has had a similar structure for some time. Gemini Business users have had this for a while too, which makes the consumer gap feel even more noticeable.

The more interesting question is whether Google will keep Projects and Notebooks as two separate systems or eventually merge them. Having two overlapping “persistent context” features in the same app is confusing, and the current documentation split between Enterprise and consumer already signals that Google has not fully resolved this internally.

For now, if you use Gemini on desktop and have been frustrated by losing context between sessions, this is the feature to watch. Keep an eye on our AI news coverage for rollout updates. And if you are evaluating which AI assistant best fits a specific business workflow, the presence or absence of persistent project workspaces is now a meaningful point of comparison, not just a cosmetic preference.

What to do about it

1. Check whether you have access to the Projects tab in your Gemini desktop app. It is in testing, so eligibility is limited and may roll out gradually.
2. If you use Gemini Business or Enterprise, review Google’s current Projects documentation to understand the knowledge boundary and file attachment options already available to you.
3. In the meantime, use Gems (file-backed custom instruction sets) as a partial workaround for persistent context in the consumer app.
4. If you are comparing AI assistants for a team deployment, factor in project management features alongside raw model performance.

Persistent context is table stakes for serious AI assistant use. When Projects lands publicly, test it against your actual workflows before treating it as a replacement for more structured workflow automation.

Source: Bing News · Google Gemini