UnSearch
Feature

Zero-Retention Mode: Privacy-First AI Search

How UnSearch's zero-retention mode ensures your queries are never stored, logged, or tracked. Built for teams that take data privacy seriously.

What is Zero-Retention Mode?

When zero-retention mode is enabled, UnSearch processes your search queries in memory and returns results without writing anything to disk or database. No query logs, no result caches, no analytics — nothing persists after the response is sent.

Why It Matters

  • **Regulatory Compliance** — Meet GDPR, HIPAA, and SOC 2 requirements
  • **Sensitive Queries** — Legal research, medical queries, financial analysis
  • **Client Confidentiality** — Law firms, consulting, healthcare
  • **Competitive Intelligence** — Keep your research private

How to Enable

Add a single header to your API requests:

X-Zero-Retention: true

Or set it globally in the SDK:

from unsearch import UnSearch client = UnSearch(api_key="key", zero_retention=True)

What Gets Retained (Nothing)

With zero-retention enabled: - Query text is not logged - Search results are not cached - No usage analytics beyond aggregate counters - IP addresses are not recorded - No data is shared with third parties

Self-Hosting for Maximum Privacy

For the highest level of privacy, self-host UnSearch on your own infrastructure. Combined with zero-retention mode, your queries never leave your network. Deploy with Docker in under 5 minutes and maintain complete control over your search data.