MCP Server Features Pricing Docs Blog FAQ Glossary About Contact Get Started
anonym.legal  ·  cloak.business  ·  Round-Trip PII

MCP Server
Round-Trip PII Protection

Prompts are anonymized before the AI sees them. Responses are deanonymized before you read them. Your workflow runs unchanged — the protection is invisible.

HOW THE ROUND-TRIP WORKS

  • Out → Detects 285+ PII types, replaces each with a reversible token before the AI sees the prompt
  • ← In Restores every original value in the AI's response automatically — you always read real names
  • 6 operators: hash names, encrypt cards, mask phones, redact SSNs — per entity type
  • E2E mode: only your client holds the mapping — server never stores real values
Quick Install Tool Reference Operators Guide
7
MCP Tools
6
Operators
26
Entity Groups
48
Languages
285+
Entity Types
On This Page

How It Works

Transparent proxy — AI never sees real PII

💻

Your AI Tool

Claude Desktop
Cursor · VS Code
Windsurf · Continue

prompt with PII
🛡️

MCP Server

Detects PII
Replaces with tokens
Restores on return

safe prompt
🤖

AI Model

Claude · GPT
Never sees
real PII data

BEFORE What you send
"Debug this for user john@acme.com
API key: sk-abc123xyz
SSN: 123-45-6789"
AFTER What AI receives
"Debug this for user <EMAIL_1>
API key: <API_KEY_1>
SSN: <US_SSN_1>"

↩ Response comes back with all tokens automatically restored to original values

Supported Integrations

stdio for Claude Desktop · HTTP for everything else

🤖

Claude Desktop

Native stdio transport — lowest latency. Full support for all 7 tools.

stdio native
⚡

Cursor IDE

HTTP transport via anonym.legal endpoint. AI-assisted coding with full PII protection.

HTTP available
🔷

VS Code

Works with GitHub Copilot Chat MCP support and MCP extension plugins.

HTTP available
🌊

Windsurf

Codeium's Windsurf IDE supports HTTP MCP servers natively.

HTTP available
🔗

Continue / Cline

Open-source coding assistants with full HTTP MCP protocol support.

HTTP available
🔌

Custom / REST API

Any tool that supports HTTP MCP transport, or direct REST API calls.

Available

Quick Install

Up and running in under 5 minutes

Add to claude_desktop_config.json:

JSON 
{
  "mcpServers": {
    "anonymize": {
      "command": "npx",
      "args": ["-y", "@anthropic-ai/mcp-server-anonym-legal"],
      "env": {
        "ANONYM_LEGAL_API_KEY": "your-api-key-here"
      }
    }
  }
}

Config file location:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Linux: ~/.config/Claude/claude_desktop_config.json

Restart Claude Desktop after saving. You should see 7 tools available in the MCP panel.

Add to Cursor MCP settings (~/.cursor/mcp.json):

JSON 
{
  "mcpServers": {
    "anonymize": {
      "url": "https://anonym.legal/mcp",
      "headers": {
        "Authorization": "Bearer your-api-key-here"
      }
    }
  }
}

HTTP transport. All 7 tools available. Cursor also supports the npx approach via CLI if you prefer.

Add to .vscode/mcp.json in your workspace (VS Code 1.99+):

JSON 
{
  "servers": {
    "anonymize": {
      "type": "http",
      "url": "https://anonym.legal/mcp",
      "headers": {
        "Authorization": "Bearer your-api-key-here"
      }
    }
  }
}

Requires VS Code 1.99+ and GitHub Copilot or MCP-compatible extension.

Any tool supporting HTTP MCP transport:

JSON 
// Generic HTTP MCP config
{
  "mcpServers": {
    "anonymize": {
      "url": "https://anonym.legal/mcp",
      "headers": {
        "Authorization": "Bearer your-api-key-here"
      }
    }
  }
}

Works with Windsurf, Continue, Cline, Zed, Msty, and any tool implementing MCP HTTP transport.

Get Started — Pick Your Product

MCP Server is available on both products. Free tier on both. See the comparison ↓

anonym.legal — 7 tools cloak.business — 10 tools

Two Products. One Technology.

Both built on the same round-trip MCP engine — choose by use case

DEVELOPER-FOCUSED
🔬

anonym.legal

anonym.legal/mcp

  • ✓ 7 MCP tools — analyze, anonymize, deanonymize, presets, sessions
  • ✓ 285+ entity types, 48 languages, 26 groups
  • ✓ 6 operators: hash · encrypt · mask · redact · replace · keep
  • ✓ Entity groups & named presets — per-project config
  • ✓ Session persistence: 24h ephemeral · 30-day vault
  • ✓ NPM: @anthropic-ai/mcp-server-anonym-legal

Best for: individual developers, Claude Desktop, Cursor, VS Code

Get Started Free →
FULL PLATFORM
🏢

cloak.business

cloak.business/mcp

  • ✓ 10 MCP tools — all of anonym.legal + batch & image
  • ✓ 320+ entity types, 70+ countries, 48 languages
  • ✓ Batch processing: 1–100 texts in a single tool call
  • ✓ Image redaction: OCR in 38 languages + visual PII blurring
  • ✓ Also includes: Chrome Extension · Office Add-in
  • ✓ NPM: cloak-business-mcp-server

Best for: teams, pipelines processing documents/images, enterprise

Get Started Free →
DETERMINISTIC
317 pattern recognizers — not probabilistic AI
ROUND-TRIP
Anonymizes out · deanonymizes in — full cycle
CHEAP
Token-based pricing, €0 free tier on both
TRUSTED
ISO 27001:2022 · German servers · Presidio

Step-by-step guides for your specific tool

Each platform has different config files, transport options, and quirks — especially VS Code, which uses "servers" instead of "mcpServers". Our platform guides cover the exact steps, troubleshooting, and advanced config for each tool.

7 MCP Tools

Complete parameter reference — ★ free · ● tokens charged

anonym_legal_analyze_text ★ free

Detect PII entities in text without modifying it. Returns entity types, positions, and confidence scores. Supports presets, entity groups, and custom ad-hoc recognizers.

ParameterTypeRequiredDescription
textstringrequiredText to analyze. Max 100,000 characters.
languagestringoptionalLanguage code. 48 supported (en, de, fr, es, it, pt, nl, pl, ru, zh, ja, ko, ar, hi…). Default: en
entitiesstring[]optionalSpecific entity types to detect, e.g. ["PERSON", "EMAIL_ADDRESS"]. Default: all types. Max 50.
entity_groupsstring[]optionalEntity groups to detect. Expands to all members: UNIVERSAL, FINANCIAL, DACH, NORTH_AMERICA, HEALTHCARE, etc. (26 groups total). Max 20.
output"full" | "summary" | "counts"optionalfull — entity text + positions + scores. summary — entity values only. counts — just entity type counts (fastest). Default: full
score_thresholdnumberoptionalMinimum confidence score 0.0–1.0. Lower = more detections, more false positives. Default: 0.5
preset"default" | "personal" | "private"optionalUse a saved preset configuration. default = standard entities, personal = your custom preset, private = private preset.
ad_hoc_recognizersobject[]optionalCustom regex patterns for domain-specific entities. Max 10. Each has: entity_type, patterns (regex + score), context (optional boost words), languages (optional).
response_format"json" | "markdown"optionalOutput format. json for structured data, markdown for human-readable. Default: json
Example
// Analyze for EU financial entities
{
  "text": "Invoice for Maria Müller, IBAN DE89370400440532013000",
  "entity_groups": ["UNIVERSAL", "FINANCIAL", "DACH"],
  "language": "de",
  "output": "summary"
}
anonym_legal_anonymize_text ● tokens charged

Anonymize text by detecting and replacing PII. Supports 6 operators per entity type. Use mode="tokenize" for reversible replacement (returns session_id) or mode="redact" for permanent removal.

ParameterTypeRequiredDescription
textstringrequiredText to anonymize. Max 100,000 characters.
mode"redact" | "tokenize"optionaltokenize — replace with reversible tokens (returns session_id). redact — permanent removal. Default: tokenize
operatorsobjectoptionalPer-entity-type operators. Key = entity type, value = operator config. See Operators section below.
entitiesstring[]optionalSpecific entity types to anonymize. Default: all detected.
entity_groupsstring[]optionalEntity groups to anonymize (expands to members). Max 20.
languagestringoptionalLanguage code. Default: en
score_thresholdnumberoptionalMinimum confidence 0.0–1.0. Default: 0.5
presetstringoptionalUse saved preset: default, personal, private.
persistence"session" | "persistent"optionalsession = 24h token lifetime. persistent = 30 days. Default: session
sessionNamestringoptionalOptional name for the session (alphanumeric, underscore, hyphen). Max 100 chars.
output"full" | "metadata" | "minimal"optionalfull = anonymized text + entity details. metadata = text + type counts. minimal = text only (fastest). Default: full
e2e_modebooleanoptionalEnd-to-end encryption mode. Server returns original text positions; client builds token mappings locally. Server never sees PII values. Default: false
response_format"json" | "markdown"optionalOutput format. Default: json
Example — per-entity operators
{
  "text": "Patient John Smith (DOB: 1985-03-15, SSN: 123-45-6789) prescribed aspirin",
  "language": "en",
  "entity_groups": ["UNIVERSAL", "NORTH_AMERICA", "HEALTHCARE"],
  "operators": {
    "PERSON": {"type": "hash", "hash_type": "SHA256"},
    "DATE_TIME": {"type": "replace", "new_value": "[DATE]"},
    "US_SSN": {"type": "mask", "chars_to_mask": 7, "from_end": false}
  },
  "persistence": "persistent",
  "sessionName": "patient-record-2026"
}
anonym_legal_detokenize_text ★ free

Restore original PII values from tokens. Pass the anonymized text and the session_id returned by anonymize_text. Works for both session (24h) and persistent (30d) tokens.

ParameterTypeRequiredDescription
textstringrequiredTokenized text containing tokens like <EMAIL_1>, <PERSON_1> to restore.
session_idstringrequiredSession ID returned by anonymize_text. Links tokens to their original values.
Example
{
  "text": "Here is the analysis for <PERSON_1> with account <EMAIL_1>",
  "session_id": "sess_abc123xyz"
}
anonym_legal_estimate_cost ★ free

Estimate token cost before processing. Use this before large anonymization runs to plan token budget. Does not modify text or charge tokens.

ParameterTypeRequiredDescription
textstringrequiredText to estimate cost for.
languagestringoptionalLanguage code. Affects token counting. Default: en
anonym_legal_get_balance ★ free

Returns your current token balance, plan tier, and usage statistics. No parameters required.

anonym_legal_list_sessions ★ free

List all active tokenization sessions. Sessions expire: session type in 24h, persistent in 30 days. Use this to find old session IDs or audit active sessions.

No required parameters. Returns an array of session objects with id, name, entity counts, persistence type, and expiry timestamps.

anonym_legal_delete_session ★ free

Permanently delete a tokenization session and all its token mappings. Use for GDPR right-to-erasure compliance or when sensitive session data is no longer needed.

ParameterTypeRequiredDescription
session_idstringrequiredSession ID to permanently delete.

6 Operators

Configure per-entity-type anonymization method via the operators parameter

replace

Replace entity with custom text. If no new_value provided, uses a type-appropriate synthetic placeholder (e.g., <PERSON>).

"PERSON": {"type": "replace", "new_value": "[REDACTED NAME]"}
// Input:  "Contact John Smith about this"
// Output: "Contact [REDACTED NAME] about this"
redact

Remove the entity entirely. No parameters. Permanent — no session created. Best for strict compliance where values must not exist in any form.

"EMAIL_ADDRESS": {"type": "redact"}
// Input:  "Send to john@acme.com for review"
// Output: "Send to  for review"
hash

Replace with cryptographic hash. Same input always produces same hash — useful for deduplication without storing originals. Choose SHA256 (64 chars) or SHA512 (128 chars).

"PERSON": {"type": "hash", "hash_type": "SHA256"}
// Input:  "Patient: John Smith"
// Output: "Patient: a8f5f167f44f4964e6c998dee827110c..."
encrypt

Encrypt with AES-256. Requires a 16, 24, or 32-character key you provide. The key is yours — anonym.legal never stores it. Decryptable by anyone with the key.

"CREDIT_CARD": {"type": "encrypt", "key": "your-32char-encryption-key-here"}
// Input:  "Card: 4532015112830366"
// Output: "Card: ENC:base64encoded..."
mask

Partially hide the value with masking characters. Set chars_to_mask for how many characters to obscure. Use from_end: false (default) to mask from start, true to mask from end.

"PHONE_NUMBER": {"type": "mask", "chars_to_mask": 6, "masking_char": "*", "from_end": true}
// Input:  "+1-555-867-5309"
// Output: "+1-555-******"
keep

Exclude an entity from anonymization. Useful when you want to detect but not anonymize certain types — e.g., keep DATE_TIME visible while anonymizing names and IDs.

"DATE_TIME": {"type": "keep"}
"PERSON": {"type": "redact"}
// Input:  "John filed on March 15, 2026"
// Output: "[PERSON] filed on March 15, 2026"

26 Entity Groups

Use entity_groups to target entire regions or domains at once

UNIVERSAL
PERSON · EMAIL_ADDRESS · PHONE_NUMBER · LOCATION · DATE_TIME · AGE · URL · IP_ADDRESS · MAC_ADDRESS · DOMAIN_NAME
FINANCIAL
CREDIT_CARD · IBAN_CODE · SWIFT_CODE · CRYPTO · UK_SORT_CODE · IN_UPI_ID
HEALTHCARE
UK_NHS · US_DEA_NUMBER · US_MEDICARE · MEDICAL_LICENSE · ICD_CODE · NDC_CODE · MEDICAL_RECORD_NUMBER · PRESCRIPTION_NUMBER · PATIENT_ID · DE_HEALTH_INSURANCE · AU_MEDICARE · NZ_NHI · IT_HEALTH_CARD · JP_HEALTH_INSURANCE
CORPORATE
DE_HANDELSREGISTER · FR_SIRET · FR_SIREN · UK_COMPANY_NUMBER · UK_VAT · ES_CIF · NL_KVK · US_EIN · JP_CORPORATE_NUMBER · IN_GSTIN · KR_BUSINESS_NUMBER · CH_UID
NORTH_AMERICA
US_SSN · US_PASSPORT · US_DRIVER_LICENSE · US_ITIN · US_BANK_NUMBER · US_DEA_NUMBER · US_EIN · US_MEDICARE · CA_SIN · CA_PASSPORT · CA_DRIVER_LICENSE
DACH
DE_TAX_ID · DE_PASSPORT · DE_DRIVER_LICENSE · DE_ID_CARD · DE_SVN · DE_HEALTH_INSURANCE · AT_TAX_ID · AT_SVN · AT_PASSPORT · CH_AHV · CH_PASSPORT · LI_ID_CARD…
UK_IRELAND
UK_NHS · UK_NINO · UK_PASSPORT · UK_DRIVER_LICENSE · UK_UTR · UK_VAT · UK_COMPANY_NUMBER · IE_PPS · IE_PASSPORT · IE_DRIVER_LICENSE · IE_VAT
FRANCE
FR_NIR · FR_CNI · FR_PASSPORT · FR_DRIVER_LICENSE · FR_TAX_ID · FR_SIRET · FR_SIREN · BE_NATIONAL_NUMBER · BE_VAT · LU_NATIONAL_ID · MC_ID_CARD…
LATIN_AMERICA
ES_NIF · ES_NIE · ES_NSS · ES_CIF · MX_CURP · MX_RFC · AR_DNI · AR_CUIT · CL_RUT · CO_CC · CO_NIT…
ASIA_PACIFIC
JP_MY_NUMBER · CN_RESIDENT_ID · HK_ID · TW_ID · SG_NRIC_FIN · SG_UEN · KR_RRN · IN_PAN · IN_AADHAAR · IN_VOTER_ID…
NORDIC
SE_PERSONNUMMER · DK_CPR · NO_FODSELSNUMMER · FI_HETU · IS_KENNITALA · SE/DK/NO/FI/IS passports & driver licenses
OCEANIA
AU_TFN · AU_ABN · AU_ACN · AU_MEDICARE · AU_PASSPORT · NZ_IRD · NZ_PASSPORT · NZ_NHI · NZ_DRIVER_LICENSE
EASTERN_EUROPE
RU_INN · RU_SNILS · RU_PASSPORT_INTERNAL · UA_IPN · UA_PASSPORT · BY_PASSPORT · BY_TAX_ID · MD_IDNP…
CENTRAL_EUROPE
CZ_RODNE_CISLO · SK_RODNE_CISLO · HU_TAX_NUMBER · HU_SSN · RO_CNP · BG_EGN · passports & IDs…
BALKANS
HR_OIB · SI_EMSO · MK_EMBG · GR_AFM · GR_AMKA · RS_JMBG · BA_JMBG · AL_NIPT · ME_JMBG · XK_ID_CARD…
BALTIC
LT_ASMENS_KODAS · LV_PERSONAS_KODS · EE_IK · passports & driver licenses for LT/LV/EE
NETHERLANDS
NL_BSN · NL_PASSPORT · NL_DRIVER_LICENSE · NL_VAT · NL_KVK
ITALY
IT_FISCAL_CODE · IT_VAT_CODE · IT_DRIVER_LICENSE · IT_PASSPORT · IT_HEALTH_CARD · IT_ID_CARD · IT_LICENSE_PLATE
LUSOPHONE
PT_NIF · PT_CC · PT_NISS · BR_CPF · BR_CNPJ · BR_RG · BR_LICENSE_PLATE · passports…
POLAND
PL_PESEL · PL_NIP · PL_REGON · PL_PASSPORT · PL_ID_CARD · PL_DRIVER_LICENSE
SOUTHERN_EUROPE
MT_ID_CARD · MT_VAT · CY_ID_CARD · CY_VAT · CY_TAX_ID · AD_NRT · SM_ID_CARD · VA_PASSPORT
MIDDLE_EAST
AE_EMIRATES_ID · SA_NATIONAL_ID · IL_ID_NUMBER · TR_KIMLIK_NO · TR_TAX_ID · passports…
VEHICLE
VIN · DE_LICENSE_PLATE · FR_LICENSE_PLATE · ES_LICENSE_PLATE · IT_LICENSE_PLATE · BR_LICENSE_PLATE
LEGAL
CASE_NUMBER · BAR_NUMBER
INSURANCE
INSURANCE_POLICY_NUMBER · DE_HEALTH_INSURANCE
EDUCATION
STUDENT_ID · EMPLOYEE_ID

Advanced Configuration

Presets, custom recognizers, e2e mode

Presets

Presets save a configuration of entity types, operators, and score thresholds. Create presets in your anonym.legal account dashboard and reference them by name.

JSON
// Use a saved preset (entities + operators all pre-configured)
{
  "text": "Your text here...",
  "preset": "personal"
}

// Preset types:
// "default"  — standard entity set
// "personal" — your saved custom preset
// "private"  — private preset with additional controls

Custom Recognizers (Ad-hoc)

Define custom regex-based entity types for domain-specific data — employee IDs, internal project codes, proprietary identifiers. Up to 10 per request.

JSON
{
  "text": "Employee EMP-2026-00142 accessed project P-ALPHA-7 on 2026-03-14",
  "ad_hoc_recognizers": [
    {
      "entity_type": "EMPLOYEE_ID",
      "patterns": [{"regex": "EMP-\\d{4}-\\d{5}", "score": 0.9}],
      "context": ["employee", "staff", "worker"]
    },
    {
      "entity_type": "PROJECT_CODE",
      "patterns": [{"regex": "P-[A-Z]+-\\d+", "score": 0.85}],
      "context": ["project", "initiative"]
    }
  ]
}

End-to-End Encryption Mode

With e2e_mode: true, the server returns original text positions instead of managing token mappings server-side. Your client builds the mapping locally and can encrypt it with its own key. The server never stores or sees the PII values.

JSON
{
  "text": "Refer patient Jane Doe (DOB: 1990-01-15) to cardiology",
  "e2e_mode": true,
  "mode": "tokenize",
  "entity_groups": ["UNIVERSAL", "HEALTHCARE"]
}
// Returns: anonymized_text + e2e_token_positions[]
// e2e_token_positions: [{token, entity_type, original_start, original_end, score}]
// Client extracts: text.slice(original_start, original_end) = "Jane Doe"
// Client stores this mapping encrypted locally — server never sees it

Use Cases

Real-world scenarios — entity groups and operators for each domain

Legal & Law Firms

AI-assisted contract review without exposing client identity

Lawyers can use Claude for contract analysis, legal research, and drafting while client names, addresses, case numbers, and financial terms stay protected. The LEGAL group plus UNIVERSAL and FINANCIAL covers most legal document types.

{
  "entity_groups": ["UNIVERSAL", "FINANCIAL", "LEGAL"],
  "operators": {
    "PERSON": {"type": "replace", "new_value": "[CLIENT]"},
    "CASE_NUMBER": {"type": "hash", "hash_type": "SHA256"},
    "IBAN_CODE": {"type": "mask", "chars_to_mask": 14, "from_end": true}
  },
  "persistence": "persistent"
}
Healthcare & Clinical

HIPAA-safe AI analysis of patient records

Clinical staff can leverage AI for diagnosis assistance and record summarization while PHI (Protected Health Information) is automatically stripped. The HEALTHCARE group covers medical record numbers, prescription numbers, patient IDs, insurance IDs, and healthcare-specific identifiers across 5 countries.

{
  "entity_groups": ["UNIVERSAL", "HEALTHCARE", "NORTH_AMERICA"],
  "operators": {
    "PERSON": {"type": "hash", "hash_type": "SHA256"},
    "DATE_TIME": {"type": "replace", "new_value": "[DATE]"},
    "US_SSN": {"type": "redact"},
    "MEDICAL_RECORD_NUMBER": {"type": "hash", "hash_type": "SHA256"}
  }
}
FinTech & Banking

AI-powered fraud analysis without exposing account data

Financial institutions can use AI for transaction analysis, fraud detection, and customer support while IBANs, credit cards, sort codes, crypto wallets, and customer identifiers are handled per compliance policy.

{
  "entity_groups": ["UNIVERSAL", "FINANCIAL"],
  "operators": {
    "CREDIT_CARD": {"type": "mask", "chars_to_mask": 12, "from_end": false},
    "IBAN_CODE": {"type": "encrypt", "key": "your-32char-key-here------------"},
    "CRYPTO": {"type": "hash", "hash_type": "SHA256"}
  }
}
Software Development & DevOps

AI code review without leaking credentials and secrets

Developers can paste code, logs, and configuration files into Claude and Cursor for AI assistance without exposing API keys, connection strings, JWT tokens, database credentials, or internal hostnames. Works automatically in the background — zero workflow change.

// The MCP Server automatically detects these in your code:
// API_KEY, JWT_TOKEN, DATABASE_URL, OAUTH_TOKEN
// AWS_SECRET, PRIVATE_KEY, CLIENT_SECRET, SSH_KEY
// IP_ADDRESS, HOSTNAME, DOMAIN_NAME, FILE_PATH

// Example: these in a paste get automatically intercepted
DATABASE_URL=postgres://user:pass@db.internal/prod
API_KEY=sk-proj-1234abcxyz
JWT_SECRET=eyJhbGciOiJIUzI1NiJ9.secret...

Limits & Rates

100K

Max text length

100,000 characters per request (approx. 80 pages of text)

100/min

Rate limit

100 requests per minute per API key

24h / 30d

Session persistence

Session tokens: 24 hours. Persistent tokens: 30 days.

The MCP Privacy Gap

Adding MCP servers to your AI tool creates a data flow you may not have considered

What happens when you add any MCP server

When Claude Desktop or Cursor calls a third-party MCP server — for database queries, file access, web search — the AI can instruct that server to process any data in the current conversation context. Privacy protections from Anthropic or Cursor apply to the model, not to MCP servers you add.

The anonym.legal MCP server intercepts every prompt before the AI sees it — including before any other MCP tool call is made. PII is replaced with tokens at the protocol layer, so no other server ever receives real personal data from your conversations.

Scenario Without anonymize.dev With anonymize.dev
AI model sees your prompt Real PII included Tokens only
Other MCP servers in context Receive real PII Receive tokens
Prompt injection targeting PII Can extract real data Only tokens extractable
GDPR Art. 25 compliance Manual — your responsibility Automated at protocol level
Developer workflow change required None needed None needed — transparent

HOW THE CHAIN WORKS

Your prompt (with PII) → anonymize_text (tokens replace PII) → AI model (sees only tokens) → other MCP tools (receive only tokens) → detokenize_text (real values restored in response)

MCP Server Pricing

MCP Server uses your token balance. Analyze operations are free. Anonymize operations charge tokens proportional to text length and entity count.

Pro

€15/mo

4,000 tokens/cycle

Top-ups: +300 tokens for €1

Get Pro

Business

€29/mo

10,000 tokens/cycle

Top-ups: +350 tokens for €1

Get Business

Note: Currently all plans include all Pro and Business features during promo period. All 7 MCP tools available. Free plan (200 tokens/cycle) also available for testing.

Pricing above is for anonym.legal. cloak.business has separate plans with batch & image tools.

View complete pricing with all plan features →

Related Resources

⚙️

Operators Deep Dive

When to use replace, hash, encrypt, mask, redact, or keep — with real-world examples for each use case.
📖

Full Documentation

REST API reference, code examples in Python, Node.js, and cURL, plus SDK documentation.
❓

FAQ

Common questions about MCP Server integration, entity detection, session management, and security.