How to Create a Team Research Workspace with Perplexity Spaces: Step-by-Step Guide
How to Create a Team Research Workspace with Perplexity Spaces
Perplexity Spaces transforms how teams conduct collaborative research by providing shared AI-powered workspaces. This step-by-step tutorial walks you through setting up project-based collections, pinning trusted sources, and configuring share links so your entire team stays aligned on research outcomes.
What Are Perplexity Spaces?
Perplexity Spaces are persistent, shareable research environments within Perplexity AI. Unlike standard single-thread searches, Spaces let you organize queries, pin reference sources, and invite collaborators — turning ad-hoc searches into structured knowledge bases for ongoing projects.
Prerequisites
- A Perplexity account (free tier works; Pro unlocks advanced features)- Team members with Perplexity accounts for collaboration- A clear project structure or research topic in mind
Step 1: Create Your First Space
- Log in to perplexity.ai and click Spaces in the left sidebar.- Click + Create Space in the top-right corner.- Enter a descriptive name for your Space (e.g.,
Q3 Competitor AnalysisorProduct Launch Research).- Add an optional description to clarify the scope and goals of the research workspace.- Select an AI model. Pro users can choose GPT-4o, Claude, or Sonar Large. Free users get the default model.- Click Create to launch your Space.
Naming Convention Best Practices
| Pattern | Example | Use Case |
|---|---|---|
| [Team] - [Project] | Marketing - SEO Audit 2026 | Department-level projects |
| [Client] - [Phase] | Acme Corp - Discovery | Client-facing research |
| [Topic] - [Quarter] | AI Regulations - Q3 2026 | Ongoing monitoring |
Pricing Models Comparison instead of your raw query).Organize threads by sub-topic to create a logical research hierarchy. For example, a *Product Launch Research* Space might contain threads for Market Sizing, Competitor Features, Regulatory Requirements, and User Interview Summaries.
Step 3: Pin Trusted Sources
Source pinning ensures that Perplexity prioritizes specific domains and documents when answering queries within your Space.
- Inside your Space, click the Settings (gear icon) or navigate to the Space configuration panel.- Find the Custom Instructions or Sources section.- Add trusted source URLs that should be prioritized. For example:
Pinned Sources: - https://arxiv.org
- https://docs.your-company.com
- https://sec.gov/edgar
https://your-internal-wiki.notion.site- You can also upload files directly to a Space (Pro feature). Supported formats include PDF, TXT, and CSV. The AI will reference uploaded documents when answering queries.- Add Custom Instructions to guide the AI’s behavior across the entire Space:Custom Instructions Example:- Always cite primary research papers over blog posts
- Focus on data from 2024-2026
- Use metric units for all measurements
Flag any information that contradicts our internal documentation
Step 4: Configure Sharing and Collaboration
- Open your Space and click the Share button (or the collaborator icon).- Choose your sharing method:
- For email invitations, enter each collaborator’s email address and select their permission level.- For shareable links, toggle the link sharing on and copy the generated URL.- Share the link via your team’s communication channel (Slack, Teams, email).Method Access Level Best For Invite by Email Full collaboration (search, view, edit threads) Core team members Shareable Link View-only or edit access depending on settings Stakeholders and reviewers Private (default) Only the creator Personal research
Step 5: Integrate with the Perplexity API (For Developers)
If your team wants to programmatically interact with Perplexity or build custom workflows, use the Perplexity API:
# Install the API client
pip install openai
Query Perplexity API with source focus
import openai
client = openai.OpenAI(
api_key=“YOUR_API_KEY”,
base_url=“https://api.perplexity.ai”
)
response = client.chat.completions.create(
model=“sonar-pro”,
messages=[
{
“role”: “system”,
“content”: “You are a research assistant. Prioritize academic sources and cite all claims.”
},
{
“role”: “user”,
“content”: “What are the latest developments in transformer architecture efficiency?”
}
]
)
print(response.choices[0].message.content)
Citations are returned in response.citations when available# CLI alternative using curl
# CLI alternative using curl
curl -X POST https://api.perplexity.ai/chat/completions
-H “Authorization: Bearer YOUR_API_KEY”
-H “Content-Type: application/json”
-d ’{
“model”: “sonar-pro”,
“messages”: [
{“role”: “system”, “content”: “Research assistant focused on cited, factual answers.”},
{“role”: “user”, “content”: “Compare React and Vue adoption rates in 2026”}
]
}‘
Pro Tips for Power Users
- Use Focus Modes per Thread: Within a Space, switch between Writing, Academic, Math, or Web focus modes depending on the thread’s purpose. Academic mode is ideal for literature reviews; Web mode works best for market research.- Layer Custom Instructions: Combine Space-level instructions with per-query system prompts. Space instructions set the baseline; individual queries can override for specific needs.- Export and Archive: Regularly export valuable threads as Markdown or PDF for offline reference. This protects against accidental deletion and creates an audit trail.- Create Template Spaces: Build a master Space with your standard instructions, source pins, and structure, then duplicate it for each new project to maintain consistency.- Leverage File Uploads: Upload meeting transcripts, research papers, or datasets directly to your Space. The AI will cross-reference uploaded content with web results for comprehensive answers.- Use the Collections Pattern: Create separate Spaces for different research phases (Discovery, Analysis, Synthesis) and cross-reference findings between them.
Troubleshooting Common Issues
| Issue | Cause | Solution |
|---|---|---|
| Collaborator cannot access Space | Email mismatch or account not created | Verify the collaborator has a Perplexity account with the exact invited email address |
| Pinned sources not being referenced | Sources may not contain relevant content for the query | Refine your query to explicitly mention the source domain, or add clearer custom instructions |
| File upload fails | File exceeds size limit or unsupported format | Ensure files are under 25MB and in PDF, TXT, or CSV format. Pro subscription required |
| API returns 401 Unauthorized | Invalid or expired API key | Regenerate your API key at perplexity.ai/settings/api |
| Threads appear out of order | Default sorting is by last activity | Rename threads with numbered prefixes (e.g., 01 - Market Size) for manual ordering |
| Model selection unavailable | Free tier limitation | Upgrade to Perplexity Pro for model selection and advanced features |
How many collaborators can I add to a single Perplexity Space?
Perplexity Spaces currently support multiple collaborators per workspace. Free accounts may have limits on the number of invitations, while Pro accounts can add more team members. All collaborators must have their own Perplexity accounts. For large teams, consider creating role-based Spaces (e.g., one for analysts, one for leadership review) rather than adding everyone to a single Space.
Can I use Perplexity Spaces with private or internal data securely?
Yes, with caveats. Files uploaded to Spaces are processed by Perplexity’s servers. For sensitive data, review Perplexity’s data handling and privacy policies before uploading confidential documents. Enterprise plans offer enhanced data privacy controls. As a best practice, avoid uploading PII or trade secrets and instead use custom instructions to direct the AI toward publicly available information about your domain.
What is the difference between Perplexity Collections and Spaces?
Perplexity evolved its organizational features over time. Spaces replaced the earlier Collections feature and offer a more robust experience. Spaces include all Collection capabilities — thread grouping and source pinning — plus additional features like file uploads, custom AI instructions, model selection, and granular sharing permissions. If you have existing Collections, they may have been migrated to Spaces automatically.