The Kaomoji Generator browses a curated collection of 97 Unicode text-face entries in ten mood categories. Search intersects with the selected mood, and Random selects and tries to copy one current result. It can repeat. The selected complete face remains available for manual copying.
97 stored entries across ten mood categories. Search and the selected category work together.
Random selects and tries to copy one current result. It can repeat a previous choice.
If Copy fails, select the complete value above and copy manually.
This is a curated library of 97 Japanese-style Unicode text faces in ten mood categories. Browse a stored expression, search its keywords and copy the complete string. It does not generate an original drawing, interpret a message or decorate arbitrary source text.
Browsing by mood, searching, copying
Start with All moods, or choose one category such as happy, angry or misc.
Search a literal face or keywords such as table flip or bear. Search only examines entries in the selected category.
Use a named Copy button, or Random from results to choose and try copying one visible result.
Check Selected kaomoji and wait for confirmation. If clipboard access fails, select the complete value and copy manually. Clear all filters restores the full collection and clears the selection.
How search matches faces
Search trims outer whitespace and matches English mood/keyword text without case sensitivity. Multiword queries require every word to appear in the combined metadata. Literal face matching is case-sensitive; for example, the uppercase O in the stored surprised face matters. These are substring matches, not an interpretation of your mood.
Category and Search intersect. Searching bear while happy is selected gives no result because the bear entry is in misc. Random uses this same result list, can repeat and is disabled when no entries match. It selects a stored face rather than assembling new parts.
Copy preserves the original sequence, including spaces, backslashes, combining marks and fullwidth punctuation. Unicode text can contain characters from several scripts. A face is not a downloaded image, but its appearance still depends on font coverage, fallback, shaping and wrapping in the receiving app. Emoji are also Unicode character sequences; they are not simply image files.
Exact stored faces and the searches that find them
Exact stored strings and the filters that find them
Category and Search
Stored face
What to check
Angry; table flip
(╯°□°）╯︵ ┻━┻
Fullwidth closing parenthesis and a space before the table
Confused; shrug
¯\_(ツ)_/¯
The backslash survives Copy
Sleepy; sleeping
(-_-)zzz
The trailing letters are part of the face
Surprised; literal face
(o_O)
Lowercase o and uppercase O are distinct
Mood browsing, literal lookup and random picks
Mood browsing
Compare the expressions stored under one mood before choosing a face.
Literal lookup
Find a known expression without losing its punctuation or backslashes.
Filtered discovery
Use Random after narrowing a keyword to explore only matching entries.
Message drafting
Paste a face beside ordinary words when the conversation suits it. Check the draft before sending.
When a search finds nothing, or a field refuses the face
No results does not mean the face is unavailable everywhere. A selected mood can hide keyword matches; choose All moods and try again. A successful clipboard write does not confirm that a username, bio or message field will accept the string.
Some entries use Unicode outside ASCII. Missing glyphs, combining-mark placement and line wrapping can alter the visual result. Do not promise identical appearance across fonts, universal platform acceptance or professional suitability. Preserve the complete face when manually copying.
Keeping the message readable
Keep essential meaning in readable words, then add a face if helpful. Use a short keyword first and add a second word to narrow the list. Inspect the pasted draft on the intended device. Character counts shown here are JavaScript UTF-16 units, not a guaranteed destination limit.
