Pricing

Starter

  • Up to 1M words per month
  • 1 language supported
  • 50+ entity types supported
  • Technical support

Scale

  • Up to 100M words per month
  • 39 languages supported
  • 50+ entity types supported
  • Technical support
  • Model optimization

Pro

  • Unlimited words per month
  • 39 languages supported
  • 50+ entity types supported
  • Dedicated support engineer
  • Model optimization
  • Custom PII entities
  • Synthetic PII generation
  • GPU support
  • Custom file formats & deployments
  • Optional air-gapped deployment

Ready To Get Started?

All tiers are hosted by the customer and include monthly model updates.
Private AI is not in the business of collecting customer data! View our privacy policy.

Privacy for Non-Profits

Private AI is free for non-profits for research purposes. Get in touch

Why Us?

Private AI delivers incredible accuracy and runs on-prem to keep you firmly in control of your data. Contact us to run our benchmark toolkit and see for yourself how we stack up against our competitors.

Private AI
AWS Comprehend
Microsoft Presidio
Google DLP
Internal Solutions

Detects & redacts PII/PCI/PHI

Does not claim rights to your data

No 3rd party processing

Hits 99.5%+ accuracy*

Detects 50+ entities

Multilingual PII detection

Custom-tuned models

Contextual synthetic PII generation

Flexible to deploy on-prem or in cloud

25x lower cloud bills

No internal maintenance required

Real-time support from actual humans

* Number quoted is the number of PII words missed as a fraction of total number of words. Computed on a 268 thousand word internal test dataset, comprising data from over 50 different sources, including web scrapes, emails and ASR transcripts. Please contact us for a copy of the code used to compute these metrics or try it yourself here.

Learn more about Private AI’s use cases, methodology, how we compare against our competitors, and more:

99.5%+ Accuracy

Number quoted is the number of PII words missed as a fraction of total number of words. Computed on a 268 thousand word internal test dataset, comprising data from over 50 different sources, including web scrapes, emails and ASR transcripts.

Please contact us for a copy of the code used to compute these metrics, try it yourself here, or download our whitepaper.