SoraPrompt

Privacy notice

SoraPrompt privacy policy

We collect the minimum amount of data required to run the service. This page explains what we store, how we use it, and how you stay in control.

Information we collect

  • Account basics: email address, Google profile id, and timestamps from Supabase when you sign in.
  • Usage telemetry: counts of prompt refines, saved prompts, and plan changes to enforce quotas.
  • Billing metadata: customer id and subscription history when you upgrade through Polar.
  • Device context: extension id, browser version, and anonymised error events for debugging.

How we use data

  • Operate the service, honour plan limits, and keep your extension session synced.
  • Provide billing receipts, manage subscription status, and surface quota resets.
  • Detect abuse, secure accounts, and improve the prompt experience with aggregate metrics.
  • Respond to support requests via team@soraprompt.ai within two business days.

What we do not do

  • We do not sell or broker personal data.
  • We do not train external models on your saved prompts or model outputs.
  • We do not read OAuth tokens outside the Chrome extension bridge workflow.

Data retention

  • Account data and prompts are retained while you keep a SoraPrompt account active.
  • Usage logs roll after 18 months; aggregated counters stay for reporting.
  • Billing records are stored for at least seven years to satisfy accounting obligations.
  • Support threads are deleted 90 days after closure unless you request otherwise.

Your choices

  • Access or delete your data by emailing privacy@soraprompt.ai from the signed-in email.
  • Revoke Chrome extension permissions from chrome://extensions at any time.
  • Export saved prompts as JSON by contacting support; we reply with a download link.
  • Opt out of release notes via the notification preferences inside the dashboard.

Subprocessors

  • Supabase for auth and database hosting (USA/EU).
  • Polar for subscription billing (EU).
  • Vercel for hosting and CDN (USA/EU).
  • OpenAI for model inference, when you request a generation.

Updates

  • We update this notice when we launch new features or change vendors.
  • Material changes are announced in-product at least seven days in advance.
  • Last updated: 02 June 2024.