@desert-ant-labs/redact

On-device multilingual PII redaction for JavaScript (node and browsers). Finds names, addresses, emails, phone numbers, cards, IBANs, national IDs and more across the 24 official EU languages, fully locally: the package runs through a local WebAssembly runtime with inference via LiteRT.js (@litertjs/core).

Inference runs in the browser (LiteRT.js is a browser runtime, using an XNNPACK-accelerated CPU path by default and optional WebGPU). In plain Node the WebAssembly pipeline still loads, but the model session is unavailable, so use this in a browser (or a browser-like environment) for redaction.

npm install @desert-ant-labs/redact @litertjs/core

import { Redact } from "@desert-ant-labs/redact";

const redact = await Redact.load();            // downloads the model on demand, cached
const r = await redact.redaction("Email Anna at [email protected].");

r.redactedText;      // "Email [GIVEN_NAME_1] at [EMAIL_1]."
r.items;             // detections: label, original, placeholder, confidence, offsets
const reply = await llm(r.redactedText);       // the LLM sees only placeholders
r.restore(reply);    // originals filled back in

Redact.load() accepts:

• directory (node): an explicit model directory; files already there are used offline, otherwise the model is downloaded into it. Omit for the managed cache (~/.cache/desert-ant-models/...).
• onProgress: download progress callback, fraction in [0, 1].
• litert: bring-your-own LiteRT.js module (the @litertjs/core namespace, e.g. a bundler-managed import).
• litertWasmDir: URL/path to the LiteRT.js Wasm files (defaults to the installed package, or the jsDelivr CDN in the browser).
• accelerator: "wasm" (XNNPACK CPU, default), "webgpu", or "webnn".

The model repo and revision are pinned to the package version. @litertjs/core is an optional peer dependency.

The same model ships as a Swift package (iOS/macOS) and an Android AAR from the same repository: https://github.com/Desert-Ant-Labs/redact

License

Desert Ant Labs Source-Available License 1.0: free below 100,000 monthly active devices per platform; above that a commercial license is required ([email protected]). Full terms: https://license.desertant.ai/1.0