Perplexity Projects for Research (Formerly Spaces)
Use Perplexity Projects (formerly Spaces) to organize research threads, files, instructions, sources, and sharing without replacing literature databases.
Naming update: Perplexity now calls Spaces Projects. This guide uses the current product name while retaining “formerly Spaces” for readers following older links and tutorials; the public URL stays unchanged.
Projects are persistent, shareable Perplexity workspaces. They group threads, files, and custom AI instructions so a thesis chapter does not collide with a side project. Projects do not replace citation managers or reproducible database searches; they are an organization and context layer around Perplexity’s search tools. For the full stack map, start with Perplexity for Researchers: A Practical 2026 Guide; for plan trade-offs, see Perplexity Free vs Pro for Students and Researchers.
Open the Projects panel in Perplexity, create one workspace for a real research question, and check the plan controls shown in your account before assuming you need an upgrade.
What a Project is (and is not)
A Project is a persistent workspace inside Perplexity that can group Search conversations, Computer tasks, files, custom instructions, connected tools, and accumulated context. It lets collaborators return to the same body of work instead of forcing every topic into one undifferentiated history.
What Projects are good at:
- Separating contexts — e.g. “Qual methods reading pack” vs “Stats refresher for committee” without cross-contaminating prompts.
- Keeping files and instructions scoped — uploaded papers and project-level instructions provide relevant context for work inside that Project.
- Returning later — you can find and continue past threads within the same Project, which matters for semester-long work.
- Selecting sources deliberately — a Project can work with web search, academic papers, social threads, file attachments, and eligible connected sources.
What Projects do not replace:
- Reference managers (Zotero, EndNote, etc.) for citations and metadata hygiene.
- Reproducible discovery (Google Scholar, databases) for defensible literature sets.
- Deep document reading where you need margin notes and stable PDF workflows — pair Projects with tools covered in How to Use AI for Reading Research Papers.
If you need a head-to-head on Perplexity vs ChatGPT for research roles, use Perplexity vs ChatGPT for Research — this article stays focused on Projects.
Sharing and current plan limits
Projects are private by default. The owner can make a Project shareable and grant view or contributor access through the Share controls. Anyone invited to a Project may be able to access files uploaded to it, so treat sharing as a data-access decision rather than a cosmetic setting.
Perplexity’s current plan documentation lists plan-dependent file and collaboration caps. As of this update, it states that Pro subscribers can upload up to 50 files per Project. Check the live help page before a deadline because file-size and plan limits can change.
If a project needs more files than your account permits, split the corpus by a defensible boundary—such as chapter or review stage—or move the canonical document set to institution-approved storage. The dedicated pricing lens is Perplexity Free vs Pro for Students and Researchers (2026).
Suggested Project layouts for common research jobs
These are templates, not rules — adapt names to how you think.
| Research job | Project naming pattern | What you store there |
|---|---|---|
| Seminar course | CourseCode — Week N | Syllabus PDFs, reading prompts, weekly synthesis threads |
| Thesis chapter | Ch3 — Methods | Methods papers, advisor feedback PDFs, revision Q&A |
| Grant / IRB prep | Grant 2026 — background | RFP PDFs, boilerplate answers, risk tables from prior years |
| Lab rotation | LabX — onboarding | SOP PDFs, instrument manuals, beginner “what does this acronym mean” threads |
Keep each Project narrow enough that its thread list and files stay interpretable. If a Project feels like “everything,” split it.
Collaboration: what a shared Project means in practice
For research teams, treat shared Projects as a coordination layer, not a data repository of record:
- Good for aligning on questions and drafting a shared reading map before people dive into primary sources.
- Poor as the only place grant budgets, human-subjects data, or embargoed drafts live — institutional storage and version control still win.
If your institution restricts third-party AI, check IT / IRB guidance before uploading sensitive materials into any Project, shared or not. Perplexity says Pro and Max users can opt out of model training in account settings, while Enterprise files and searches are excluded from AI training; those controls do not replace institutional approval.
Failure modes (and how to avoid them)
-
Confusing a Project with a literature database — Projects help you navigate; they do not verify completeness of a field. When stakes go up, move the canonical paper set to Scholar / your reference manager (Perplexity vs Google Scholar).
-
One mega-Project — if every query lands in “Research,” you recreated the default thread with extra steps. Split early.
-
Treating a shared Project as private storage — collaborators can access files in a shared Project. Check membership and sharing settings before uploading anything restricted.
How Projects fit the four-stage workflow
The hub’s Perplexity for Researchers: A Practical 2026 Guide maps orientation → discovery → reading → writing. Projects are strongest in orientation and early discovery:
- Orientation: keep quick threads that map jargon and schools of thought inside one Project.
- Discovery: select web and academic-paper sources, then move the verified candidate set into your reference manager.
- Reading / writing: use uploaded files and custom instructions for bounded questions, but do not let the Project become the only place your evidence or argument lives.
For stage-by-stage tool choice across the whole stack, see AI Research Workflow: Which Tool for Which Stage.
Official sources checked
- What are Projects? — current naming, workspace capabilities, and sharing controls
- Perplexity plan comparison — current plan-level feature and privacy differences
- Data Collection at Perplexity — consumer opt-out and Enterprise training policy
Related reading
- Perplexity for Researchers: A Practical 2026 Guide — parent hub: source selection, Projects, and verification habits
- Perplexity Free vs Pro for Students and Researchers (2026) — when Project, Research, and file limits can justify Pro
- Perplexity vs ChatGPT for Research — different jobs, different tools
- Best AI Tools for PhD Students and Researchers in 2026 — budget-first stack placement