Privian
Get Started

Pricing

Simple pricing for privacy-first AI workflows

Start small, scale when needed, and secure prompts before sensitive data reaches GPT, Claude and other models.

BYOK · Privacy-first · Zero raw prompt retention

Plans

Choose the right plan

  • Minimum

    $19/mo

    For solo builders and small AI projects.

    • Up to 50k requests / month
    • PII masking
    • Prompt security
    • Privacy-first gateway
    • BYOK provider support
    • Supported entity masking
    • Developer docs
    • Email support

  • Pro

    $79/mo

    For production workloads and growing teams.

    • Up to 250k requests / month
    • Everything in Minimum
    • Priority support
  • Early supporter

    Founding 100

    $299one-time

    Early supporter pricing for the first 100 customers during beta. Equivalent access to the Pro plan.

    95 founding spots remaining5/100 claimed
    • Everything in Pro
    • Future pricing grandfathered
    • Priority beta feedback access
    • Up to 250k requests / month
    • Provider-agnostic
    • BYOK support

    Early supporter pricing for teams helping shape Privian during beta. Fair-use limits apply. Limited to the first 100 customers.

  • Enterprise

    Contact

    For organizations with custom requirements.

    • Custom requirements
    • Roadmap discussions
    • Priority onboarding
    Contact us

The same core platform is included in every plan. Plans differ by monthly request volume and support level.

LLM provider usage costs (OpenAI, Anthropic and others) are billed directly by your provider through BYOK. Privian only charges for the gateway and security layer.

Compare

Plan comparison

FeatureMinimumProFounding 100Enterprise
Requests / month50k250k250kCustom
PII maskingYesYesYesYes
Prompt securityYesYesYesYes
BYOKYesYesYesYes
Supported entity maskingYesYesYesYes
Email supportYesYesYesYes
Priority supportYesYesYes
Priority beta feedbackYesYes
Grandfathered pricingYesCustom
Custom onboardingYes

FAQ

Pricing FAQ

What's the difference between the plans?
The same core platform is included in every plan. Plans differ by monthly request volume and support level. Founding 100 adds an early-supporter pricing relationship — grandfathered pricing and priority beta feedback — with access equivalent to Pro.
Why BYOK?
Privian is bring-your-own-key. You connect your own OpenAI, Anthropic or other provider account, and Privian routes prompts through its security layer. You keep direct ownership of provider billing, quotas, and data agreements.
Do requests include provider usage costs?
No. Privian charges only for the AI security layer. Provider token usage (OpenAI, Anthropic, etc.) is billed by your provider directly on your own account.
What happens after beta?
Pricing and plan limits may change as Privian leaves beta. Founding 100 customers keep their pricing — see grandfathered pricing below.
What is Founding 100?
Founding 100 is a one-time early supporter plan for the first 100 customers. It includes Pro-equivalent access (same features, same 250k requests/month cap), priority beta feedback, and grandfathered pricing on future plans. Fair-use limits apply.
What does grandfathered pricing mean?
If general pricing increases after beta, Founding 100 customers continue at their original terms for the equivalent feature set. We will document the exact mapping when general pricing is finalized.
Does Privian store prompts?
No. Privian operates with zero raw prompt retention. Prompts are processed in memory through the security layer and discarded after the request completes. See the Security page for the full data-handling model.
Can I upgrade later?
Yes. You can move between plans at any time. Founding 100 access is preserved for customers who joined during the beta program.

Transparency

Built transparently during beta

Privian is evolving in public with implementation-grounded documentation, transparent limitations and privacy-first defaults.

  • BYOK provider model

    You bring your own OpenAI, Anthropic or other provider keys. Privian never owns provider billing or quotas.

  • Zero raw prompt retention

    Prompts and rehydrated responses are not persisted. Observability is built from counters and latencies.

  • Documented masking types

    Supported entity types are listed in the docs so you can verify coverage before integrating.

  • Transparent beta limitations

    What's not yet supported is published on the Updates page — no compliance theater.

  • Public architecture and security model

    Architecture, data handling and the request lifecycle are documented openly.

  • Privacy-first defaults

    Masking is on by default. The model only ever sees the masked prompt.

See our architecture, security and updates pages.

Ready to secure your AI workflows?

Start with Minimum, or join the Founding 100 to lock in early supporter pricing.

Get StartedTalk to us