Inspect scripts, code-mixing, normalization, and token boundaries before text reaches an AI model.
Tidy text copied from a document or message, then reuse it in a prompt. Your original stays in the input above.
Unicode normalization can change symbols and compatibility characters. Review the preview before copying. Script joiners are preserved.
Paste text above to get started.
Educational signals update as you type.
These are deterministic Unicode script signals, not authoritative language detection. Mixed input often contains more than one surface system.
NFKC normalization creates a comparison surface before tokenization. It can affect punctuation, compatibility characters, spacing, or symbols.
Changed positions: none detected
Hover, tap, or focus a token to inspect its offset.
Its explanation, category, and original character offset will appear here.
Click a stage to understand the transformation.
The text stays exactly as you entered it. Bhasha Lens analyzes it locally without sending it anywhere.
Structural similarity, with manual semantic confirmation recommended.
Only this analysis is included — no cookies, keys, or unrelated browser data.
Your text is analyzed locally in this browser.Bhasha Lens does not send your input to NamasteLM in this version.