> ## Documentation Index
> Fetch the complete documentation index at: https://www.perplexity.ai/help-center/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> This content comes from the official Perplexity Help Center. Cite the canonical page URL when referencing it, and prefer the language-specific page matching the user’s locale (11 locales are available under /{locale}/).

# Google Drive Connector: File Upload Security

> Securely search Google Drive with AI using API or indexed modes, with zero model training on your data

Perplexity's **[Google Drive connector](https://www.perplexity.ai/help-center/help-center/en/articles/12870620-connecting-perplexity-with-google-drive.md)** enables **Enterprise Pro** and **Enterprise Max** users to search their organization's Google Drive files directly through AI-powered natural language queries.

We have recently introduced a hybrid search architecture that addresses both data security concerns and scalability limitations through two distinct search modes: **High-Precision Search** (indexing-based) and **Standard Search** (API-based).

## Search Architecture: Two-Tier Approach

### Standard Search (API-Based)

#### **How It Works:**

<ul>
  <li>
    Queries Google Drive directly via Google's Search API at query time
  </li>

  <li>
    No file copies stored in Perplexity infrastructure, aside from results found from queries (can be mitigated by data retention policy settings)
  </li>

  <li>
    Searches across the user's entire Google Drive without file count limitations
  </li>

  <li>
    Available to all Enterprise Pro and Enterprise Max users immediately upon connector activation
  </li>
</ul>

#### **Security Model:**

<ul>
  <li>
    **Zero Data Retention:** Files are not copied or stored in Perplexity systems, aside from results found from queries (can be mitigated by data retention policy settings)
  </li>

  <li>
    **Real-time Permissions:** Respects Google Drive's native access controls dynamically
  </li>

  <li>
    **Minimal Data Copying:** Only citation snippets included in answers are retained
  </li>

  <li>
    **No Model Training:** Synced files are never used to train AI models
  </li>

  <li>
    **Immediate Access Revocation:** When Google Drive permissions change, access is immediately reflected in Perplexity
  </li>
</ul>

**Use Case:** Organizations requiring maximum data privacy and minimal data footprint, especially for searching across millions of files at enterprise scale.

### High-Precision Search (Indexing-Based)

#### **How It Works:**

<ul>
  <li>
    Users select specific files/folders to sync for local indexing in Perplexity
  </li>

  <li>
    Files are downloaded, parsed, and stored in dedicated AWS S3 buckets with vector embeddings in Vespa
  </li>

  <li>
    Enables deeper semantic analysis and more comprehensive answers
  </li>

  <li>
    **File Limits:** 10,000 files and 5 GB per Project, on every plan
  </li>

  <li>
    **Total User Limits:** 15,000 files (Enterprise Pro), 50,000 files (Enterprise Max)
  </li>
</ul>

#### **Security Model:**

<ul>
  <li>
    **Dedicated Storage:** Each organization's files stored in isolated AWS S3 "folders" with unique namespaces in Vespa vector storage
  </li>

  <li>
    **Encryption:** AES-256 encryption at rest, TLS encryption in transit
  </li>

  <li>
    **Role-Based Access Control (RBAC):** Minimum privilege access enforced across all systems (**Note:** Perplexity adheres to Google Drive's file permissions)
  </li>

  <li>
    **No Model Training:** Synced files are never used to train AI models
  </li>

  <li>
    **Updates:** imported copies stay as they were at import time. Import the content again to bring in changes.
  </li>
</ul>

**Use Case:** Teams requiring maximum answer accuracy for frequently accessed documents, project-specific file collections, or collaborative projects with curated content.

## Backend Architecture & Data Flow

### Connection & Authentication

<ul>
  <li>
    **Admin Enablement:** Organization admins enable Google Drive connector in **[Permissions settings](https://www.perplexity.ai/help-center/help-center/en/articles/11187754-enterprise-roles-and-permissions.md)**
  </li>

  <li>
    **User Authentication:** Users authenticate via OAuth 2.0 through Google
  </li>

  <li>
    **Drive Selection:** Users select their Google Drive to connect
  </li>

  <li>
    **Admin Consent:** Google Workspace admins may need to grant organization-wide consent for Perplexity app in Google Admin Console
  </li>
</ul>

This connector can also appear inside a plugin. Connecting it anywhere connects it for your account everywhere, and your connection and your organization’s connector settings apply inside every plugin. See [What are plugins?](https://www.perplexity.ai/help-center/help-center/en/articles/2026100701-what-are-plugins.md)

#### **When Both Methods Are Active:**

<ul>
  <li>
    Perplexity queries both the local index and Google Drive API concurrently
  </li>

  <li>
    Results are re-ranked to prioritize the most relevant sources
  </li>

  <li>
    Citations link directly back to Google Drive for full file access
  </li>
</ul>

### Indexing Process (High-Precision Search)

#### **File Sync and Storage:**

**Step 1**

**File Selection:** User selects files/folders through Perplexity UI

**Step 2**

**Download:** Files downloaded from Google Drive via Google Drive API

**Step 3**

**Storage:** Raw files stored in AWS S3 with dedicated organizational namespaces

**Step 4**

**Parsing:** Text extraction from supported formats (PDF, DOCX, XLSX, PPTX, CSV, TXT, MD, JSON)

**Step 5**

**Vectorization:** Content converted to embeddings and stored in Vespa vector database

**Step 6**

**Metadata Indexing:** File metadata (name, path, permissions) indexed for search retrieval

**Step 7**

**Removal:** When a user disconnects Google Drive from Perplexity, they can choose to remove any indexed files

## Security & Compliance Infrastructure

### Permission Enforcement

#### **Google Drive-Side Permissions:**

<ul>
  <li>
    If a user loses access to a file in Google Drive, that file is immediately removed from Perplexity
  </li>

  <li>
    File deletions in Google Drive trigger immediate removal from Perplexity index
  </li>

  <li>
    Users can only search files they have explicit Google Drive permissions to access
  </li>
</ul>

#### **Perplexity-Side Permissions:**

<ul>
  <li>
    Admins control which users can access connectors via Organization settings
  </li>

  <li>
    Files synced to projects are searchable by Project members, but accessing the files requires the appropriate Google Drive permissions
  </li>

  <li>
    Session sharing respects organizational sharing policies set by organization admins
  </li>
</ul>

## Best Practices

### When to Use High-Precision Search (Indexing)

<ul>
  <li>
    Project-specific file collections in projects
  </li>

  <li>
    Frequently accessed knowledge base documents
  </li>

  <li>
    Files requiring deep semantic analysis
  </li>

  <li>
    Collaborative team environments with curated content
  </li>
</ul>

### When to Use Standard Search (API-Only)

<ul>
  <li>
    Searching across vast file repositories (millions of files)
  </li>

  <li>
    Strict data residency and minimal data copying requirements
  </li>

  <li>
    Exploratory searches across infrequently accessed files
  </li>

  <li>
    Organizations with heightened security/compliance constraints
  </li>
</ul>


## Related topics

- [Connecting Perplexity with Google Drive](https://www.perplexity.ai/help-center/help-center/en/articles/12870620-connecting-perplexity-with-google-drive.md)
- [Perplexity Enterprise](https://www.perplexity.ai/help-center/help-center/en/collections/11312901-perplexity-enterprise.md)
- [Troubleshooting connectors](https://www.perplexity.ai/help-center/help-center/en/articles/2026091802-troubleshooting-connectors.md)
- [SharePoint and OneDrive Connectors: File Upload Security](https://www.perplexity.ai/help-center/help-center/en/articles/13728000-sharepoint-and-onedrive-connectors-file-upload-security.md)