ShinrAI uses semantic encryption to replace sensitive details with realistic alternatives before they reach your AI provider. The text stays useful. Your application controls the mapping to restore the originals.
14-day money-back guarantee · Cancel monthly · Invoice with VAT
Wherever people write about people, ShinrAI lets AI help without exposure.
Protect patient documents, doctor letters and clinical archives on your own infrastructure before using them in AI analysis.
Process documents separately or in batches, keeping replacement mappings under your control.
Integration guidance →Help advisors and analysts work with customer correspondence while replacing personal details and supported financial identifiers.
Integrate protection into customer records, correspondence and internal AI workflows.
Explore this workflow →Prepare citizen correspondence, case files and administrative records for AI processing within your chosen deployment.
Operational usage records track consumption without storing submitted document text.
Integration guidance →Draft, summarise and review matters that name real people while parties, addresses and firms travel as consistent, restorable stand-ins.
Keep control of the mapping so authorised applications can restore the original details.
See an employee workflow →Build protection into customer-facing apps, websites and services using the API and documented integrations.
API keys authenticate your application. AI connections let an assistant use tools with your permission.
Integration guidance →Summarise tickets, draft replies and analyse feedback while replacing customer details before they enter an AI prompt.
Protect retrieved CRM records, emails and indexed documents as part of your workflow.
See a customer workflow →Let your assistant look up a customer request. Choose which sources it retrieves before preparing a reply.
Protect documents before indexing, and protect newly retrieved text before it enters an AI prompt. Your application keeps permissions and the original mapping under its control.
Every classic tool can blank out a name. ShinrAI can too — but a blanked text loses the information the AI needs to reason. Multi-dimensional replacement keeps it.
Dear [PERSON], your appointment at [ORGANIZATION] in [CITY] on [DATE] is confirmed. Please bring [PERSON]'s referral from [PERSON].
Dear Paul Brandt, your appointment at Klinikum Nordpark in Biberach on May 12 is confirmed. Please bring Anna Vogel's referral from Dr. Weber.
| Original | Detected realm | mask | preserve | Why not the “obvious” swap? |
|---|---|---|---|---|
Robert |
PERSON · DE · common · masc · given | [FIRSTNAME] |
Paul |
Never “Mohammed” — flipping cultural origin measurably changes AI answers. |
Bad Neustadt |
CITY · DE · small | [CITY] |
Biberach |
Never “München” — a small-town→metropolis jump breaks the story’s logic. |
Mailand (in German text) |
CITY · IT · major | [CITY] |
Turin |
The entity is Italian even when the text is German — exonym-aware, culture kept. |
Aleksandra Nowak |
PERSON · PL · uncommon · fem · full | [PERSON] |
Kasia Wiśniewska |
Gender is never flipped; rarity tier is never dropped. |
Siemens |
ORG · DE · international | [COMPANY] |
Bosch |
An international brand stays an international brand — scale is meaning. |
Autohaus Krüger GmbH |
ORG · DE · regional | [COMPANY] |
Fuhrmann Fahrzeugtechnik GmbH |
Fictional but plausible — never a real firm, never a mismatched size. |
Worked examples from the ShinrAI replacement specification. Three policy modes — mask, preserve (default) and neutralize — are configurable per entity type and per deployment.
The policy modes describe the replacement specification. Available API controls depend on your deployment; check the current API documentation. API documentation
Compare the same strict-span locale tests across versions. Choose a language to see the change.
pl+1.8 strict-span F1 points from ShinrAI 1.4 to 1.5.
Partial F1 on the 4-head REDACT-PII view.
Strict-span F1 mean on ai4privacy and MultiNERD.
Locales · detection heads · entity classes. New: dates and ages. Hebrew remains beta.
Different tests answer different questions. The 15-locale median is 95.5 for 1.5 (1.4: 95.4). Review the full board—including methods, absolute scores and limitations—and pilot on your own documents.
Find your product. Follow a short guide. Bring ShinrAI into your everyday workflow.
Private AI chat, your models and ShinrAI protection in one workspace. Start with the built-in privacy controls.
Use with Secure AI Suite →
Create protected document copies and protect selected Assistant tasks.
Installation package and guide
Protect questions and company knowledge before model inference.
Installation package and guide
Protect selected model calls and knowledge uploads.
Installation package and guide
Protect contact form data and attachments before exporting them.
Installation package and guide
Protect chats, retrieved context and new document uploads.
Installation package and guide
Connect your software to the ShinrAI API. We run the service on STACKIT infrastructure in Germany; you manage your usage and billing in Dashboard.
Estimate managed usage →Run ShinrAI close to your applications on your own hardware. For disconnected deployments, import a signed activation file and keep processing local.
Explore on-premises →Start with the work your team does. Adjust the assumptions to estimate your monthly managed API usage.
Track orders, explain returns and draft replies using customer and order context.
Illustrative volume: 100.000 customer conversations per month.
Help customers with accounts, subscriptions and service issues, with a handover to your support team when needed.
Illustrative volume: 500.000 customer conversations per month.
One record covers up to 1.000 characters with Standard processing. Each text is rounded up separately.
A monthly allowance for regular usage.
14-day money-back guarantee on your first month
Not the right fit? Write to support within 14 days of your first plan payment and we refund it in full. Once per company, if you used no more than 10 % of the plan's monthly records.
For small law firms and tax offices protecting everyday AI work and an internal chatbot.
For growing SMEs using AI across customer records, support tickets, sales and internal workflows.
For teams building AI into customer-facing apps, websites and services.
For larger organisations and software providers running AI across multiple products, departments or customers.
Top up an existing plan or start with a standalone pack. Pack records do not expire and are used after the current plan allowance.
€0,318 per 1.000 records
€0,182 per 1.000 records
€0,147 per 1.000 records
Prices exclude applicable tax. Checkout calculates the correct business treatment, and each purchase is followed by an invoice email.
Choose a plan or pack above to continue.
Change your selection →14-day money-back guarantee · Cancel monthly · Invoice with VAT
No account needed. After payment we email you a secure sign-in link and your first API key.
Already have an account? Open Dashboard.
Run detection and replacement in your own environment. For disconnected deployments, a signed activation file binds usage to your installation; processing stays local.
Get help integrating the API, designing data flows and training your team. Choose the managed service or an on-premises deployment that fits your environment.
Meet our partners →These packs activate local usage on your own infrastructure. They are separate from managed API records and bind to one installation. The container can process requests without an internet connection after activation.
Based on 5,000 characters per submitted document, processed once at Standard tier.
Approximately 1 million documents
Standard and Batch
Approximately 5 million documents
Standard, Batch and Real-Time
Approximately 20 million documents
Standard, Batch and Real-Time
Prices exclude applicable tax. Activation and file download are available after payment confirmation in Dashboard.
Reversible replacement is pseudonymisation and does not by itself establish GDPR or EU AI Act compliance.
GDPR principles · European Commission ↗ · EU AI Act · European Commission ↗
Compare ShinrAI with Azure PII and other privacy tools, from restoration and hosting to pricing and measured quality.
The previous model, ShinrAI 1.3, is available as open weights on Hugging Face. Licence and download →
Our certified partners help you choose the right setup, connect your workflows and support your team as usage grows.
Competence Center of Excellence for software packaging, deployment and automation.
Cloud migration, identity and endpoint management.
Automate repetitive business processes and connect workflows.
Extract, classify and route information from documents.
Help businesses put private AI to work.
Become a ShinrAI partner →
EECC develops RFID and traceability solutions that connect physical products with digital information. Founded by GS1 Germany, DHL and METRO, it brings applied research and software expertise to ShinrAI’s technology development.
Visit EECC ↗
Forschungszentrum Jülich advances scientific research through high-performance computing and AI. Its Jülich Supercomputing Centre operates JUPITER, ranked fifth worldwide in the June 2026 TOP500 list.
FZ Jülich ↗Include an API request ID if you have one. Keep API keys, passwords and personal data from your documents out of your message.
Get started quickly, understand metering and explore the API reference.
Browse the docs →