Wubi (五笔86) input-code lookup. Type a Chinese character to see its Wubi86 code (simple + full code), or type a Wubi code to reverse-find matching characters. Browser-only, nothing uploaded.
Wubi (86) Input Code Lookup (五筆)
How to use
Pick a lookup direction
Two modes sit at the top: "Char → code" finds the Wubi code for a character; "Code → chars" reverse-looks-up characters from a Wubi code. Click to switch.
Character → code
Type one or more characters (a whole sentence works) and press Look Up. Each character shows its Wubi86 code, highlighting the shortest "simple code" and the full code.
Code → characters
Switch to "Code → chars", type a Wubi code made of letters a–z (e.g. k or khk), and press Look Up to list the characters that code produces. Exact and prefix matches are supported.
Read the result
All codes are Wubi86 (五笔字型 86 edition). The data comes from the open-source Rime Wubi table; everything runs locally in your browser — nothing is uploaded or stored.
Wubi: A Shape-Based Chinese Input Method
Wubi (五笔字型, "five-stroke") is the most widely used shape-based Chinese input method on the mainland, invented by Wang Yongmin in 1983. Unlike pinyin input, which encodes a character by how it sounds, Wubi encodes a character by how it is built — every character is decomposed into a sequence of "radicals" (字根) that are laid out across the 25 letter keys a–z (z is the learning / wildcard key). Those keys are grouped into five regions following a character's opening stroke — horizontal, vertical, left-falling, right-falling, and turning — which is exactly where the name "five-stroke" comes from. Because of this, Wubi lets you type a character without knowing its pronunciation, which makes it especially friendly for rare characters, technical text, and users not fluent in Mandarin, while avoiding the heavy homophone clash that plagues pinyin input.
Simple codes vs full codes
The code this tool returns for a character is often not just one string. Wubi assigns simple codes to high-frequency characters to speed up typing: a level-1 simple code needs just a single keypress (the 25 most common characters each own one key, e.g. 中 = k, 我 = q, 人 = w), a level-2 simple code needs two keys, and a level-3 simple code needs three; the full code (usually four letters) is the standard code produced by the complete decomposition rules. If a character has a simple code, fluent typists reach for it first, but the full code works in every implementation. In the results, the shorter code is the simple code and the longer one is the full code — the tool labels each so you can memorise whichever you need.
Reverse lookup: from code to character
The tool's second mode is a reverse lookup — type a string of Wubi letters and the tool searches a reverse index it builds on load to find every character that code produces. This is handy for learning Wubi, checking a radical decomposition, or cracking "which character does this letter string make?" Note that one code can map to several candidate characters (depending on the homophone/collision order); the tool lists exact matches first, then prefix-related characters. Every match is a deterministic table lookup — no model, no randomness is involved.
"Wubi typing ignores sound and reads shape — if you can take a character apart, you can type it."
About this tool's data
This tool uses the Wubi86 table (五笔字型 86 edition), sourced from the open-source Rime input-method Wubi table. Other editions exist — Wubi98, New Century, and so on — with slightly different radical regions and individual codes, so the same character can carry a different code across editions; if you use the 98 edition, defer to the table that ships with your own IME. This tool is for lookup, learning, and cross-reference only. All computation runs locally in your browser — nothing is uploaded or stored.
10 Facts about Wubi Input
Wubi was invented by Wang Yongmin in 1983 — the first practical "shape-code" Chinese input scheme on the mainland, once hailed as a key technology that "brought Chinese characters into the computer".
The "five strokes" are the five basic strokes of Chinese characters: horizontal (一), vertical (丨), left-falling (丿), right-falling (丶) and turning (乙). The keys a–z are split into five regions on this basis, five keys each.
Wubi encodes by shape, not sound, so in principle any character can be typed — including rare and variant characters. This is its biggest advantage over pinyin.
Fluent Wubi users hit very few collisions — most characters appear on the first try with little candidate-paging — so blind-typing speed can be very high, long making it the choice of professional typists.
Level-1 simple codes: the 25 most common characters each own one letter key and are typed with a single press — e.g. 中 = k, 我 = q, 人 = w, 了 = b.
A full code is usually four letters: a character takes at most four radical codes; when it has fewer radicals, rules such as the "last-stroke identifier" pad it to the standard length.
Wubi has several editions: the classic is "Wubi86", followed by "Wubi98" and a "New Century" edition. Their radical regions and some codes differ, so they are not fully compatible.
The z key is Wubi's "learning / wildcard" key: in most implementations it matches any unknown radical, helping beginners look up the correct code while they type.
As pinyin input and predictive typing spread, the share of Wubi users has fallen — but it retains a steady following in professional data entry, classical-text editing, and legal and medical documents.
Because Wubi codes are unrelated to sound, homophones no longer interfere: 在 / 再 / 载 are told apart by shape, instead of crowding into one "zai" and forcing repeated candidate selection as in pinyin.
This tool uses the open-source Rime Wubi86 table and runs entirely in your browser — nothing is uploaded or stored — so you can safely batch-check codes with confidence.
Frequently Asked Questions
-
Wubi is a shape-based input method that encodes a character by its form rather than its sound, invented by Wang Yongmin in 1983. It decomposes each character into radicals mapped onto the a–z keys, so you can type without knowing the pronunciation. This tool looks up the Wubi86 code for any character, or reverse-finds characters from a code.
-
It uses the Wubi86 (五笔字型 86 edition) table, taken from the open-source Rime Wubi table. Wubi98 and a New Century edition also exist, with different radical regions and some different codes. If you use the 98 edition, codes may differ slightly — defer to the table shipped with your IME.
-
Because Wubi gives high-frequency characters simple codes. The shorter string is the simple code (a level-1 simple code needs just one key, e.g. 中 = k); the longer one (usually four letters) is the full code. Fluent users prefer the simple code to save keystrokes, but the full code works everywhere. The tool lists every code for a character and labels which is which.
-
A simple code is a shortened code for common characters: level-1 takes one key, level-2 two, level-3 three. A full code is the standard code from the complete decomposition rules, usually four letters. If a character has a simple code it is faster to type, but every implementation accepts the full code.
-
In "Code → chars" mode, the tool builds a reverse index (code → character) when the data loads; you type a string of Wubi letters and it finds the matching characters. Exact matches are shown first, followed by characters whose code starts with what you typed. One code can map to several candidates.
-
No. Apart from loading this tool's own code-table JSON once, there are no network requests and no AI / large language model involved. Every lookup is a deterministic local table query: the same input always returns the same result.
-
No. All lookups happen locally in your browser with no server calls, no tracking, and no storage. The characters or codes you type never leave your device.
-
Two main reasons: (1) different edition — your IME may be Wubi98 or New Century, whose radical regions differ from the 86 edition this tool uses; (2) different simple-code choices — implementations don't always assign the same simple-code level to a character. This tool follows the open-source Rime Wubi86 table.
-
Yes. Wubi decomposes by shape, so traditional characters also have codes — though the radical breakdown may differ from the simplified form. This tool's table is primarily simplified, with some traditional/variant characters included; if a traditional character returns no code, try its simplified counterpart.
-
Each has strengths; there is no absolute answer. Wubi has few collisions, fast blind typing, handles rare characters, and ignores pronunciation — but it has a steeper learning curve, requiring you to memorise radicals and decomposition rules. Pinyin has almost no learning cost and is universal, but suffers many collisions and frequent candidate selection. Professional entry and classical-text work lean on Wubi; everyday typing leans on pinyin. This tool just looks up codes — it doesn't choose for you.
Related News
You may be interested in these recent stories from our newsroom.
-
AWS Commits $1 Billion to Embedded AI Engineers as the Enterprise Fight Shifts to Deployment
AWS has committed US$1 billion to a Forward Deployed Engineering organisation that embeds its engineers inside customer teams, following pri...
-
ByteDance Launches Seedream 5.0 Pro, an Image Model That Outputs Editable Layers
ByteDance has launched Seedream 5.0 Pro, a professional image-generation and editing model that breaks a single render into more than ten ed...
-
Mira Murati's Thinking Machines Releases Inkling, Its First Open-Weight Model
Mira Murati's Thinking Machines has released Inkling, its first model: an open-weight, multimodal Mixture-of-Experts system with 975 billion...