Instagram Hidden Words is a native way to filter unwanted language in comments and message requests. It includes platform protections and lets account owners add custom words, phrases and emojis. Keep it on if it helps your account. SABHYA is a separate Instagram comment moderation product by The Indian Alpha. In its private beta, it provides a review-oriented workflow for decisions that need more context; it is not a replacement for Instagram’s controls.
What Hidden Words can do
Meta says Hidden Words can filter potentially offensive language and supports custom lists of words, phrases and emojis. Meta has also described improvements for some misspellings and additional languages. Availability and settings may change, so check the controls shown in your Instagram account and Meta’s explanation of Hidden Words. Instagram also provides other native safety tools, including Limits for periods of unwanted attention.
| Need | Useful first step | Where human review helps |
|---|---|---|
| Known unwanted terms | Use Instagram’s built-in filters and a carefully maintained custom list. | Check whether a filtered word appeared as a quotation or support message. |
| Repeated targeting | Use account controls and platform reporting as appropriate. | Review the pattern and preserve relevant evidence. |
| Unclear slang or mixed-language text | Avoid treating a single term as a complete verdict. | Examine the thread, intended target and creator policy. |
These approaches are complementary. The table is guidance for a moderation process, not a test of either system.
Why context can change a decision
The following examples are synthetic and do not show SABHYA’s measured performance:
- “That insult was aimed at me in yesterday’s comments” quotes harmful language but may be a request for help.
- “Yeh idea kaafi useful tha” mixes Romanized Hindi and English but expresses praise.
- A phrase that looks mild in isolation may become targeted harassment in a reply thread.
Romanized Hindi spelling varies, and Hinglish often switches languages inside one sentence. A language label or word match may help triage, but neither establishes intent by itself. Read the Hinglish moderation guide for the research and practical limits.
Where SABHYA fits
SABHYA distinguishes KEEP, REVIEW and HIDE as moderation decisions. REVIEW is for uncertainty that should go to a person; a HIDE decision is separate from confirmation that Instagram performed an action. The private beta starts with human review, and automatic moderation is still being validated. The workflow explainer and system card describe its scope.
Limitations
Neither a custom word list nor contextual assistance can guarantee that every harmful comment is caught or every legitimate comment is preserved. Creator policy, platform controls and human judgment remain important. SABHYA does not publish a live Instagram accuracy benchmark here.
Frequently asked questions
Should I turn off Hidden Words to use SABHYA?
No. Continue using the native settings that work for your account. SABHYA is designed as a complementary review workflow.
Can Hidden Words handle Hinglish?
Meta has described language expansion and misspelling improvements, but coverage can vary. Test settings with your own account and review mistakes; do not assume any language or spelling is perfectly covered.
Is SABHYA publicly available?
SABHYA is in controlled private beta for its initial Instagram Professional account scope. You can request beta access or explore the product.
Related reading
See the Instagram comment moderation guide and the native-controls evidence comparison.