Cruxy
  • Enterprise
  • Pricing
  • Docs
ContactGet started
Get started

Products

  • All products
  • Cruxy
  • Memory
  • Vaults
  • Artifacts
  • Cruxy Code
  • Moderation API
  • Cruxy Cowork
  • Cruxy Bench
  • Cruxy Guard

Models

  • Auto
  • Fast · Mira
  • Balanced · Vaani
  • Deep · Kavi

Developers

  • Overview
  • Docs
  • API reference
  • Model reference
  • API pricing
  • Status

Pricing

  • All plans
  • Free
  • Core
  • Advanced
  • Pro
  • Enterprise

Company

  • About
  • Contact sales
  • Sign in
  • Report a vulnerability

Legal

  • Security
  • Trust
  • Privacy
  • Terms
  • Usage policy
  • Cookies

Cruxy. The AI that works with you.

[email protected]© 2026 Mycrux Private Limited
Language: English
  • EnglishEnglish
  • हिन्दीHindi
  • EspañolSpanish
  • FrançaisFrench
  • DeutschGerman
  • Português (Brasil)Portuguese (Brazil)
  • ItalianoItalian
  • 简体中文Chinese (Simplified)
  • 日本語Japanese
  • 한국어Korean
  • РусскийRussian
  • TürkçeTurkish
  • Bahasa IndonesiaIndonesian
  • Tiếng ViệtVietnamese
  • বাংলাBengali
  • తెలుగుTelugu
  • தமிழ்Tamil
  • मराठीMarathi
  • العربيةArabic
  • اردوUrdu
  • فارسیPersian

Trust/Security

Your data. Your control.

What protects your data, in the same words as our privacy policy, and what is not in place yet, said just as plainly.

Report a vulnerabilityRead the privacy policy

In place

What protects your data.

In transit

Live
  • Every connection is encrypted with TLS.
  • Our sites send HSTS, so browsers connect to them only over HTTPS.

Secrets at rest

Live
  • Secrets are encrypted field by field with AES-256-GCM: the key the app uses to call the API for your account, two-factor (TOTP) secrets, OAuth tokens and the credentials you give us for MCP servers.
  • API keys are stored only as hashes. Passwords are hashed with argon2id.

Access

Live
  • Only authorized personnel can reach production systems, and every access is logged.
  • Secrets and credentials are kept out of source control and rotated when team members change.

Backups

Live
  • Database backups are encrypted.
  • A backup is deleted once it is more than 30 days old; the newest seven are always kept.

Updates

Live
  • Operating systems, runtimes and dependencies get regular security updates.

Where

Live
  • Our servers and databases are in Mumbai, India.
  • Some sub-processors operate globally; each one is assessed before we engage it.

NOT YET

  • Conversation content, memories and files are not encrypted at rest. Secrets are; the rest is not.
  • No compliance certification. We will say so here when one exists.

Retention

How long we keep what. And how to have it deleted.

A job that runs once a day deletes records past their period, so a record can outlast it by up to a day. You can ask for earlier deletion at any time.

Conversations
Kept until you delete them, or 365 days after their last activity. Deleting one is immediate and permanent; there is no trash.
Temporary chats
Save nothing to memory. A chat filed into a vault is always saved.
Memories
Kept until deleted. You can delete any personal memory in Settings, or all of them at once through the API.
Site analytics
Visits and page views without cookies, IP addresses or user agents, counted on our own servers and deleted after 24 months.
API usage records
Account, key, model, token counts, cost, timing and status for each request. Never the prompt or the reply.
Backups
Deleted data stays in encrypted backups until those backups are deleted, at most 30 days.
Your rights
Access, correction and erasure: email [email protected] with the subject "Data subject request". We respond within 30 days.

Sub-processors

Who processes your data. What each one receives is in the privacy policy.

What each receives
PROVIDERPURPOSELOCATION
HostingerServers and databasesMumbai, India
CloudflareCDN, DNS, file storage and encrypted backupsUnited States, global
DeepInfraAI model inferenceUnited States
Brave SearchWeb searchUnited States
RazorpayPaymentsIndia
ResendTransactional emailUnited States
Google, Microsoft, GitHubSign-inUnited States
SentryError trackingUnited States
SlackInternal notificationsUnited States
Healthchecks.ioOperational monitoringLatvia, EU

Live

For your own apps

The moderation API is live.

Classify text against content-safety categories with cruxy-guard before it reaches chat, through the Cruxy API, with any key on any plan.

Moderation guide

Cruxy GuardComing soon

Guard, the security layer for AI work, is coming soon and is not part of any plan yet. One line, and no demo until it is real.

What is planned

RESPONSIBLE DISCLOSURE

Found something?

Write to [email protected]. The same contact is published in our security.txt. If an incident affects your personal data, we notify affected users and authorities as the law requires.

YOUR RIGHTS

Access, correction, erasure.

Email [email protected] with the subject "Data subject request". We respond within 30 days, at no cost, and never hold it against you.