Spelling, grammar, punctuation, style, and clarity β color-coded inline. Accept or reject each suggestion. Free, no signup.
Up to 1,000 words on the free tier, 5,000 signed in. Essays, emails, posts, captions β anything written.
Each issue is color-coded by category: spelling, grammar, punctuation, style, or clarity. Click an underline to see the suggested fix.
Accept the suggestions you agree with, reject the rest. Or hit Apply All if you trust them. Then copy the fixed text.
Yes. 5 checks per day with no signup. Sign in (free account) for 50/day. No credit card required at any tier.
TextSight uses Claude Sonnet 4 β a large language model that understands context. Where rule-based checkers flag every long sentence or passive voice as an issue, an LLM only flags problems that actually hurt readability. Tradeoff: slightly slower (5-15s vs instant) but generally fewer false positives.
Five categories: spelling, grammar (subject-verb agreement, tense, articles), punctuation, style (wordiness, redundancy), and clarity (ambiguous references, awkward phrasing). Each is color-coded so you can quickly skim by type.
Yes. Each issue has Accept and Reject buttons. You stay in control β accept what makes sense, reject what doesn't fit your voice. There's also an Apply All button if you trust all the suggestions.
Up to 1,000 words on the free tier. 5,000 words signed in. For longer documents, check in sections.
It's optimized for English. The underlying model can handle other languages, but quality is variable. For non-English work, use the AI Translator first or check with a native speaker.
3 free scans/day, no card. See how your writing reads to AI detectors.