# Sift - Spam Text Filter — Open source spam text filter

> Open source spam text filter (iPhone/iPad app by 源鹏 廖.)

- Source: https://appshunter.io/ios/app/sift-spam-text-filter/id6788805739 (this page in markdown: same URL + `.md`)
- Developer: [源鹏 廖](https://appshunter.io/developer/6777849850)
- Category: Utilities, Lifestyle
- Price: Free with in-app purchases
- Age rating: 4+
- Requires: iOS 18.0 · 36 MB
- Languages: American English, Japanese, Chinese (Simplified, China)
- Released: 2026-07-25
- Data updated: 2026-08-18
- User reviews in markdown: https://appshunter.io/ios/app/sift-spam-text-filter/id6788805739/reviews.md

## What is Sift?

Sift is a privacy-first SMS filtering app for iPhone, built to make your inbox quieter and easier to control.

Sift uses the iOS message filtering extension to classify junk, promotions, and important notifications directly on your device. By default, filtering does not send message contents to a server. Your custom rules, local statistics, and on-device personalization stay on the device.

Key features
- On-device SMS classification for spam, promotions, verification codes, delivery updates, banking notices, travel alerts, and more
- Custom sender or message-body rules
- Category mapping that redirects selected categories to Promotions or Junk
- Private statistics that count categories without storing message text
- Optional sample contribution with an on-device redaction preview, export, and deletion controls
- Sift Premium: a one-time purchase that unlocks an on-demand multilingual Transformer model

Sift is not a fraud-prevention guarantee. SMS classification is probabilistic and can make mistakes. For payments, accounts, verification codes, or other important messages, always verify the original message and official channels.

## Pricing and in-app purchases

Base price: Free.

**One-time purchases**

| Purchase | Price | Type |
| --- | --- | --- |
| Sift Premium | $0.99 | one-time |



## Version history (last 4 releases)

### 1.3 — 2026-08-13

- Fix some already-known issues

### 1.2 — 2026-08-05

Added support for more detailed classification mapping settings and fixed several known issues.

### 1.1 — 2026-08-01

Improved the recognition accuracy of the model, and some known issues have been fixed.

### 1.0 — 2026-07-25

No release notes.

---

*Data collected daily from the US App Store and indexed by [AppsHunter](https://appshunter.io/). User reviews are verbatim App Store reviews. Ratings, prices and chart positions refresh continuously; this snapshot is from 2026-08-18.*