AI Engine
It depends on whether you are using a personal account or an enterprise/organizational plan. For enterprise, business, and organizational customers (including Enterprise Pro), Perplexity states that it acts as a data processor, meaning your prompts and responses are not used to train or fine-tune models, and your data remains controlled under your organization’s policies. This is explicitly noted in Perplexity’s privacy materials, which say that the standard Privacy Policy does not apply to Enterprise Pro or API usage because Perplexity is operating solely as a processor in those cases.
For personal consumer accounts, Perplexity’s help documentation states that some user data may be used to train or improve models and search quality unless you opt out. Perplexity provides a data-use opt-out toggle in Account Settings. If you turn this off, your chats are not included for model improvement, though Perplexity may still temporarily store certain data for operational, safety, and service-reliability reasons.
Perplexity also clarifies in its public help resources that third-party model providers used by Perplexity are not allowed to train on Perplexity’s user data, meaning your chats are not passed on to external model vendors for training, regardless of account type.
The retention period for your chats and other logged data in Perplexity depends on how long your account remains active, the plan you are on, and whether you delete your account.
For personal accounts, Perplexity saves your chat Threads in your Library for as long as your account is active, and you can delete individual Threads at any time; if you delete your account entirely, your data is removed within about 30 days. If you use Perplexity without signing in, your conversations are stored only temporarily as anonymous Threads and are automatically deleted after 14 days. Perplexity states that it retains data only as long as necessary to provide the service and maintain security.
For Enterprise / Enterprise Pro / organizational users, search/chat history retention does not follow the personal defaults. Instead, retention is controlled by the organization’s admin, who can set a fixed retention window (for example, 30 days or another defined duration). Once data passes that retention limit, the chat records automatically become inaccessible and are permanently deleted within approximately 7 days.
For personal Perplexity accounts, you can control whether your activity is used to improve models by adjusting your Data Collection and privacy settings. Perplexity states that some user data may be used to improve model quality unless you opt out, and you can also delete individual Threads from your Library while keeping your account active. Your chat Threads remain stored in your Library for as long as your account is active, unless you manually delete them. If you use Perplexity without signing in, conversations are temporary “anonymous Threads” and are automatically deleted after 14 days.
For Enterprise accounts, you do not need to opt out because Perplexity explicitly states that enterprise data is not used to train or fine-tune models. In these environments, Perplexity acts purely as a data processor, and your organization controls data retention. The admin can set a retention duration (e.g., 30 days or longer). Once Threads pass that retention window, they become inaccessible and are permanently deleted within approximately 7 days.
Perplexity handles data differently depending on whether you are using a personal account or an Enterprise/Work/Government plan.
Personal accounts may have their chat data used to help improve Perplexity’s models unless the user opts out in settings, and data retention is primarily user-managed. In contrast, Enterprise, organizational, and Government plans include stricter privacy and data-handling guarantees: Perplexity states that Enterprise data will not be used to train or fine-tune models, and Government deployments apply an enforced zero-data-usage policy for applicable traffic. These plans also provide administrator-level retention controls, where organizations can define how long search and chat history is stored; once the retention window expires, the data becomes inaccessible and is typically permanently deleted within about 7 days.
For Enterprise / Work / Organization accounts, retention and deletion are managed by the admin, who can set a fixed retention window for Threads and enforce file-handling rules. Perplexity also states that uploaded files in Enterprise plans are automatically deleted after 7 days, and admins can control access and data lifecycle through organizational settings.
For personal accounts, Perplexity provides the ability to delete individual Threads from your Library and to delete your entire account, after which your personal data is removed from Perplexity’s systems within ~30 days. However, full data export or long-term retention controls are not comprehensively documented for personal users in current public resources. If you need a full export of your chat history or copies of stored data, users can request data or deletion through account settings or by contacting support.
Yes. API usage, especially through Perplexity’s Sonar API or Enterprise API, is handled under stricter data-use and retention rules than the consumer app.
Perplexity states that the Sonar API follows a “Zero Data Retention Policy”, meaning that data sent through the API is not stored and is not used to train or improve models beyond processing the immediate request. By contrast, the consumer app version may use data to improve model quality unless you opt out in your account settings. The Privacy Policy also separates Business/Enterprise Services from personal use, indicating stronger privacy, non-training assurances, and organizational controls for API and enterprise environments.
