What: A complete guide to Perplexity Spaces — the collaborative research feature that turns AI search into a persistent, team-friendly knowledge base.
Who: Teams, researchers, and professionals who manage ongoing research projects and want organized, shareable AI-powered workspaces.
Best if: You work on multi-day or multi-person research projects and lose track of findings across separate search sessions.
Skip if: You only use Perplexity for quick one-off queries and do not need to organize or share your research.
Bottom Line Up Front
Perplexity Spaces is the feature that transforms Perplexity from a search tool into a research platform. Spaces create persistent, collaborative environments where every search builds on previous findings, team members contribute research to shared pools, and AI context accumulates over days and weeks. For anyone managing complex research projects — competitive intelligence, academic theses, investigative reporting, strategic planning — Spaces eliminate the “lost research” problem where valuable findings disappear across browser tabs and separate sessions. Free users get 3 Spaces; Pro users get unlimited. This guide covers everything from initial setup to advanced team workflows that professional research teams are using in production.
Key Takeaways
- Spaces maintain context across sessions, so returning to a Space after days or weeks preserves all accumulated knowledge
- Shared Spaces allow team members to contribute research that benefits everyone in the group
- Custom instructions let you configure each Space for specific research goals, source preferences, and output formats
- File uploads within Spaces create persistent document libraries that the AI references across all queries
- The most effective Spaces are organized around specific projects, not broad topics
What Spaces Are and Why They Matter
Without Spaces, every Perplexity conversation exists in isolation. You search, get an answer, and when you start a new search thread, all that context is gone. This is fine for one-off questions but terrible for research projects that span multiple sessions, topics, and team members.
Spaces solve this by creating dedicated research environments with persistent memory. When you search within a Space, the AI has access to all previous searches, uploaded documents, and accumulated context within that Space. Ask “How does this relate to what we found last week about competitor pricing?” and Perplexity actually knows what you found last week because it is all stored in the Space.
According to Grokipedia, Perplexity Spaces was launched in late 2024 and has since become the platform’s fastest-growing feature, with over 5 million Spaces created by March 2026. The feature draws on concepts from collaborative knowledge management tools like Notion and Confluence but adds AI-powered search and synthesis that those tools lack.
Setting Up Your First Space
Step 1: Create the Space. Click “Spaces” in the left sidebar, then “New Space.” Choose a descriptive name that will still make sense in three months. “Q2 2026 Competitor Analysis” is good. “Research” is bad.
Step 2: Write custom instructions. Every Space lets you set custom instructions that guide all searches within it. This is remarkably powerful and wildly underused. Good custom instructions include: your research goals, preferred source types, output format preferences, and any domain-specific context the AI should know. For example: “This Space tracks competitor activity in the enterprise CRM market. Focus on pricing changes, feature launches, and customer sentiment. Prioritize sources from 2025-2026. Present findings in bullet-point format with citation links.”
Step 3: Upload foundational documents. If your research builds on existing documents — previous reports, internal data, reference materials — upload them to the Space. The AI will reference these documents alongside web search results, providing answers that integrate your proprietary knowledge with current web information.
Step 4: Set your initial context. Your first search in the Space should establish context: “I am researching [topic] for [purpose]. The key questions I need to answer are [list questions]. Help me start by providing an overview of the current state of [topic].” This primes the Space with context that improves every subsequent search.
Team Collaboration in Shared Spaces
Shared Spaces are where the feature truly shines. When multiple team members contribute research to the same Space, the AI draws on everyone’s accumulated findings to provide increasingly rich and contextual answers.
Inviting collaborators: Open the Space settings and add team members by email. Each collaborator gets full access to search within the Space and view all previous searches. Their contributions become part of the Space’s knowledge base, accessible to all members.
Division of research: Assign different research areas to different team members. In a market research context, one analyst might focus on competitor products while another tracks market trends. Both contribute to the same Space, and the AI can synthesize findings across both areas: “Based on the competitor analysis and the market trends research in this Space, where are the biggest opportunities?”
Knowledge accumulation: Over weeks and months, a shared Space becomes a rich knowledge base. New team members can ask “Summarize the key findings in this Space” to get up to speed instantly. The Space essentially becomes an institutional memory that grows smarter over time. According to Stanford HAI’s collaboration research, teams using AI-powered shared research tools produce 28% more comprehensive analyses than teams using traditional document-sharing approaches.
Advanced Space Strategies
Strategy 1: Competitor Monitoring Spaces
Create one Space per key competitor. Custom instructions: “Track all news, product updates, pricing changes, hiring activity, and strategic moves by [Company X]. Present new findings in chronological order with date stamps.” Visit weekly with: “What has changed for [Company X] since my last visit?” The Space’s accumulated context ensures you get an update that builds on previous intelligence rather than starting from scratch.
Strategy 2: Academic Research Spaces
For students and researchers, create one Space per thesis chapter or research question. Upload your literature review bibliography and any key papers. Custom instructions: “This Space covers [research topic]. Prioritize peer-reviewed sources from [relevant databases]. Present findings with full academic citations.” Each research session adds to the accumulated knowledge, and you can ask cross-referencing questions that span all your previous findings.
Strategy 3: Client or Project Spaces
Consultants and agencies benefit from client-specific Spaces. Upload the client brief, background materials, and any research done so far. Custom instructions: “This Space supports consulting engagement for [Client Name] in [Industry]. Focus on [key areas]. Maintain professional language suitable for client deliverables.” Every team member who contributes research enhances the collective understanding, and the Space can generate synthesis reports drawing on all accumulated intelligence.
Managing Multiple Spaces Effectively
As your Space library grows, organization becomes important. Here are the management practices that power users follow.
Naming convention: Use a consistent format like [Category] – [Specific Topic] – [Date or Quarter]. Examples: “Market – Enterprise CRM – Q2 2026” or “Academic – Transformer Efficiency – Thesis Ch.3.” This makes Spaces findable as your library grows.
Archiving: When a project ends, rename the Space to include “[ARCHIVED]” and remove collaborators. The Space retains all its content for future reference but does not clutter your active workspace.
Regular maintenance: Monthly, review your Spaces. Archive completed projects. Update custom instructions on active Spaces to reflect evolving research needs. Delete Spaces that are no longer useful to free up your Space count (relevant for free users with the 3-Space limit).
For journalists managing beat coverage, Spaces become an essential part of the workflow — one per beat, maintained over months, accumulating context that makes every new story faster to research. According to McKinsey, professionals who organize AI research tools by project report 35% higher satisfaction and 22% better research outcomes than those who use AI tools ad hoc.
Custom Instructions: The Hidden Power Feature
Custom instructions are the most underused feature within Spaces. They let you configure each Space to behave differently based on the research context, and most users either skip them or write generic instructions that do not meaningfully improve results.
Effective custom instructions include: Your research objective in one sentence. The types of sources you prefer (academic, industry, government). Output format preferences (bullet points, detailed analysis, structured comparisons). Domain context the AI should assume (your field, your role, your expertise level). And any specific constraints (date range, geography, language).
Example for a market research Space: “I am a market analyst researching the European enterprise SaaS market. Prioritize sources from 2025-2026 including analyst reports, financial disclosures, and customer surveys. Present findings with specific numbers and cite all data sources. When comparing vendors, use structured tables. Exclude sources older than 18 months unless they contain foundational market sizing.”
Example for an academic Space: “I am a PhD candidate in computational neuroscience. Prioritize peer-reviewed papers from Nature, Science, PNAS, eLife, and arXiv. When discussing methods, include specific technical details about model architectures and training procedures. Flag when findings are from preprints that have not undergone peer review.”
Well-crafted custom instructions can improve answer relevance by 40-50% compared to Spaces without instructions, because every query in the Space inherits this context automatically. You do not need to repeat your constraints in every search — the Space remembers them for you. Update custom instructions as your research focus evolves to keep the Space aligned with your current needs.
Spaces vs Alternatives
| Feature | Perplexity Spaces | ChatGPT Projects | NotebookLM | Notion AI |
|---|---|---|---|---|
| Web Search | Full, real-time | Limited browsing | No web search | Limited |
| Citation Quality | Excellent | Minimal | Document-based | Minimal |
| Team Collaboration | Built-in sharing | No sharing | Google Workspace | Full collaboration |
| Document Upload | Yes (Pro: 50MB) | Yes | Yes (core feature) | Yes |
| Persistent Context | Yes | Yes | Yes | Limited |
| Custom Instructions | Per-Space | Per-Project | No | Global |
| Free Tier | 3 Spaces | Yes | Fully free | Limited AI queries |
Frequently Asked Questions
How many Spaces can I create on the free plan?
Free users can create up to 3 Spaces. Perplexity Pro ($20/month) provides unlimited Spaces. If you need more than 3 active research projects, Pro is the most cost-effective upgrade because Spaces is the feature that most Pro users cite as their primary reason for subscribing. You can archive old Spaces to free up slots, but archived Spaces lose their active context until unarchived.
Can collaborators see all my searches within a shared Space?
Yes. All searches conducted within a shared Space are visible to all collaborators. This transparency is intentional — it is what allows the AI to draw on collective research. If you need to conduct private searches on a related topic, do them outside the shared Space in your personal search area. Only searches you explicitly conduct within the shared Space become visible to the team.
Do uploaded documents in Spaces count toward storage limits?
Yes. Free users have a 25MB per-file limit with 3 uploads per day across all Spaces. Pro users have a 50MB per-file limit with no daily cap. For large document libraries, consider uploading only the most important reference materials directly. For supplementary documents, summarize key points in your custom instructions instead of uploading the full files.
Can I export all research from a Space?
You can share individual threads within a Space via URL. For comprehensive export, you will need to share each thread separately or use the “summarize this Space” query to generate a synthesis document you can copy. Perplexity does not currently offer bulk export of Space content, which is a frequently requested feature. For critical research, periodically export key findings to an external document as backup.
What happens to a Space if the creator’s account is deleted?
If the Space creator deletes their account, the Space and all its content are permanently deleted, even if other collaborators still have access. For team Spaces, designate the team lead or a shared organizational account as the Space creator. Enterprise plans include admin controls that prevent Space loss when individual accounts are removed, making them the recommended option for organizations that rely on Spaces for critical research.
Get Smarter About AI Every Morning
Free daily newsletter — one story, one tool, one tip. Plain English, no jargon.
Free forever. Unsubscribe anytime.
You May Also Like
- Perplexity Computer for Taxes: What It Is, How to Use It, and What to Watch Out For
- Perplexity vs Gemini for Research: Which AI Finds Better Answers?
- Perplexity for Neighborhood Research: Help Clients Find the Right Area
- Perplexity for Student Research: Find Sources That Actually Work
- How to Use Perplexity AI: Complete Beginner’s Guide 2026
Sources
This article draws on official documentation, product pages, and industry reporting. Specific sources are linked inline throughout the text.
Last reviewed: April 2026