Online Tool Store
Bộ lọc ngôn từ tục tĩu
Phát hiện tục tĩu và lời lăng mạ trong văn bản và xem trước đầu ra đã được che, xóa hoặc gắn cờ, với phân chia mức độ nghiêm trọng xử lý khác nhau giữa nhẹ và nặng.
🔒 Công cụ này chạy hoàn toàn trên trình duyệt của bạn. Tệp của bạn không bao giờ được tải lên máy chủ.
Text & Writing
Profanity Filter
Frontend preview — no upload or external service.
Filtered output
6 matches in 340 words: 2 severe (masked), 4 mild (flagged but left in place) — one match is inside a place name and marked as a likely false positive.
How the Profanity Filter works
- Paste the text, comment, or transcript you need to review.
- Choose an action and a severity threshold — masking everything makes ordinary text unreadable.
- Read the flagged false positives before accepting the output; place names and surnames are the usual culprits.
FAQ
Why does it flag ordinary words?
Because substring matching finds real words inside longer innocent ones — the classic Scunthorpe problem. Whole-word matching reduces it but cannot eliminate ambiguity, so review is part of the process.
Can a filter catch deliberate evasion?
Only partially. Spacing, character substitution, and invented spellings defeat any word list eventually. Filtering is a first pass, not moderation.
Is my text sent anywhere?
No. Matching happens in your browser, which matters when the text is user-submitted content you are responsible for.
How we compare
| Feature | Online Tool Store | A word processor | A paid writing assistant |
|---|---|---|---|
| Severity levels | ✓ | ✗ | ✓ |
| Flags likely false positives | ✓ | ✗ | Sometimes |
| Text stays local | ✓ | ✓ | ✗ |
| Catches evasive spellings | ✗ | ✗ | Partially |
Profanity Filter is honest about its limits: word lists catch the obvious, miss the creative, and always need a human to review the flags.