Private AI is now Limina: what it offers
As of 17 November 2026, www.private-ai.com answers with a permanent redirect to getlimina.ai. Limina's site describes its main product, Data De-identification, as a way to "remove personal data from text, images, audio, and documents", alongside PII, PHI and PCI detection, redaction, pseudonymization and synthetic data generation.
- Coverage. Limina lists 52 languages and more than 50 entity types across PII, PHI and PCI, for text, PDFs, images, audio and structured data.
- Deployment. A container that runs on-premises or in your VPC, or a SaaS version where data is processed in Limina's cloud. It is also offered through the AWS, Snowflake and Azure marketplaces.
- Getting started. The pricing page lists a free Starter plan of up to 75 API calls a day, with Batch and Enterprise plans quoted.
For teams that want a self-serve API across many languages, that is a real advantage, and GritWorks does not offer an equivalent.
Questions that separate the options
Most de-identification products detect similar entity types. The differences that decide an evaluation are usually elsewhere:
- Must processing be offline? Several options run in your VPC or on-premises. Ask whether they need any outbound connection (licensing, telemetry, model updates). GritRedact's offline deployment runs with no network egress and supports air-gapped environments.
- Who reviews before release? An API returns redacted output. A regulated workflow often needs an operator to accept or correct detections before permanent removal, with the decision recorded.
- What evidence do you keep? Ask for the audit record per file: detections, actions, locations, operator, policy, and input and output hashes.
- Is removal permanent in the file? For PDFs and images, check the text layer, the pixels and the metadata after redaction, not only the API response.
- What happens next to the data? If the same team needs fields extracted to JSON or synthetic test documents with known values, one vendor across redaction, extraction and generation can simplify governance. GritWorks covers all three with GritRedact, GritTransform and GritRender.
- Language coverage. If you need broad multilingual coverage out of the box, compare the vendors' published language lists against your documents.
When to choose which
- Choose Limina when you want a de-identification API you can start on a free plan, broad language coverage, and the choice of its cloud or a container.
- Choose Presidio when you want free, open-source components for mostly text data and will build the rest yourself.
- Choose Tonic Textual when your goal is AI-ready text and files with realistic, consistent replacement values, and pay-as-you-go pricing suits you.
- Choose a cloud PII service when your data may be processed in that cloud and you want a managed API there.
- Choose GritRedact when documents, scans, images and recordings must stay inside your boundary, an operator must review before permanent removal, and you need an audit trail per file. Pricing is agreed with sales.
Before deciding, run the same labelled sample through each shortlisted tool. The redaction vendor evaluation guide explains how, and the Presidio comparison covers what an open-source build involves.
What GritWorks does not replace
- GritWorks complements authorization, governance and e-discovery tools rather than replacing them. Access control, data catalogues, retention policy and legal review platforms keep their jobs; GritWorks prepares the documents, images and audio that move between them.
- Redaction is not authorization. If a person or system is not entitled to a record, block access instead of sending a redacted copy.
- Automated redaction alone is not a declaration of HIPAA de-identification. HIPAA recognizes the Safe Harbor and Expert Determination methods; policies, validation and human review must match the intended use and recipient.
- PII flagging is not redaction. GritTransform can flag potentially sensitive values in extracted results, but that does not remove them from the source. Use GritRedact when content must be removed.
- Engagement is sales-assisted. GritWorks has no public or self-serve API, SDK, MCP server or free tier, and publishes no price list: pricing is agreed with sales. Customer APIs are provided after onboarding.
Common questions
What happened to Private AI?
Private AI's website, private-ai.com, now redirects permanently to getlimina.ai, and the de-identification product is presented under the Limina name. Check Limina's own announcements for the details of the change.
What are the main alternatives to Private AI?
Open-source Presidio, Tonic Textual, managed PII services from Azure, AWS and Google Cloud, and GritRedact for on-premises redaction of documents, images and audio with human review and an audit trail.
Can Limina run on-premises?
Limina says its product deploys as a container in your on-premises environment or VPC, and also offers a SaaS version where data is processed in its cloud.
How is GritRedact different from Limina?
GritRedact is built around a reviewed workflow for documents, images and audio: detect, human review, permanent removal and an audit trail with input and output hashes, running fully offline with no network egress. Limina offers a self-serve API with a free plan and broad language coverage. GritWorks has no free tier or public API.
Is redaction enough for HIPAA de-identification?
No. Automated redaction alone is not a declaration of HIPAA de-identification. HIPAA recognizes the Safe Harbor and Expert Determination methods, and policies, validation and human review must match the intended use and recipient.
Does GritWorks replace access control or e-discovery tools?
No. GritWorks complements authorization, governance and e-discovery tools rather than replacing them. Redaction is not authorization: if someone is not entitled to a record, block access instead.