小篆 Small Seal Script Converter
Convert modern Chinese into 小篆 small seal script, the form standardised under Qin Shi Huang around 221 BCE and recorded in the 說文解字. Handles simplified and traditional input per character, shows the modern character and pinyin beside every glyph, and leaves a slip deliberately blank where no seal form is recorded rather than inventing one.
小篆 Small Seal Script Converter
延年益寿 — prolong life
How we decide. A character is shown in seal script only when the 說文解字 records a form for it and this typeface carries it. Where it does not, we show nothing rather than guess. Seal forms varied by hand, era and medium; what you see is one modern typeface's interpretation of a standardised form, not a facsimile of any particular inscription.
Seal glyphs are rendered with 崇羲篆體 (Chong Xi Small Seal), by 王心怡 and 季旭昇, published by 中央研究院 (Academia Sinica) 小學堂 under CC BY-ND 3.0 TW or later — https://xiaoxue.iis.sinica.edu.tw/chongxi/. The file is served exactly as published, unmodified. Of the 11,604 characters it can draw, this tool draws only the 10,387 recorded in the 大徐本《說文解字》. Scholarly entries (headwords, radicals and glosses) come from the shuowen.org 說文解字 dataset of 9,833 entries, released under the Apache-2.0 licence.
How to Use the Small Seal Script Converter
Type or paste your Chinese
Any modern Chinese works — simplified or traditional, a single name, a four-character idiom, or a whole line of poetry. The converter resolves every character independently, so mixed simplified-and-traditional text is handled correctly rather than being force-converted one way first.
Read the seal forms side by side
Each character appears in its 說文解字 seal form beside the modern one, so you can compare stroke for stroke and watch a radical change shape. The curves, the balance, and the vertical stacking are what make small seal script instantly recognisable.
Understand the blank slips
A blank bamboo slip means no attested seal form exists for that character in our sources. The converter will not draw a glyph it cannot source — that refusal is deliberate, and the FAQ explains the two different reasons a blank appears.
Save it, or copy it
Download the rendering as an image for a chop design, a card, or a classroom handout. Note that Copy gives you the modern Chinese text, not the seal glyphs — the seal look is a typeface, so copied text will display in whatever font the destination uses.
Small Seal Script: the Qin Standard and What Survives of It
The script that standardised an empire
Small seal script — 小篆, xiaozhuan — is the writing system the Qin state imposed on the Chinese world once it had finished conquering it. When Qin Shi Huang completed the unification around 221 BCE, the six states he had absorbed each wrote the same language in visibly different hands, and the chancellor 李斯 (Li Si) is traditionally credited with pruning those regional variants down to a single official standard. What survives of that standard reaches us mainly through a dictionary: the 說文解字 (Shuowen Jiezi), completed by the Han scholar 許慎 (Xu Shen) around 100 CE, which analysed some nine thousand characters and, crucially, recorded the seal form of each one. Almost every seal glyph anyone can look at today — including the ones this converter draws — descends from that record rather than from a stone inscription.
That distinction matters more than it sounds. This tool does not claim to show you "the Qin form" of a character. It shows the form recorded in the 說文解字, which is a Han scholar's account of a Qin standard written three centuries after the fact — and later hands added to it. The 新附 (xinfu, "newly appended") characters were inserted by 徐鉉 (Xu Xuan) in 986 CE, more than eight hundred years after Xu Shen. The Shuowen is a magnificent source, and it is the source, but it is a dictionary with a history of its own, not a photograph of a Qin bamboo slip.
Why the converter sometimes gives you nothing
Type 咖啡 (coffee), 氧 (oxygen), 硅 (silicon) or 艦 (warship) and you will get a blank bamboo slip back. Nothing is broken. There is no seal form for those characters because the things they name — the drink, the element, the semiconductor, the steel warship — did not exist when the script was standardised, and nobody in the Qin or the Han ever had occasion to write them. Simplified characters (简化字) create the same gap for a different reason: they were promulgated from 1956, more than two thousand years after Li Si, so as written shapes they were never carved in seal script at all.
Most converters paper over this. They synthesise a plausible-looking glyph, or they quietly fall back to your system font so the "seal" output ends up containing ordinary modern characters. This one refuses. Where no attested form exists you get an empty slip and a note, because a fabricated seal character is worse than no character — it teaches you something false about the history of the script. Measured on a corpus of 400 four-character idioms, 93.5% of individual characters resolved to an attested seal form and 77.3% of the idioms resolved completely. Those are honest numbers, and the idioms that come back partial are the point rather than the failure.
采 (to pick) has a seal form. Its traditional elaboration 採 does not — 采 is the older graph. Convert your text to traditional first and you can destroy coverage instead of improving it.
Simplified, traditional, and one character at a time
The 采 case is not a curiosity; it is why the converter works character by character. A document-level 简 to 繁 pass assumes traditional forms are always the older, better-attested ones. They are not. Sometimes the simplified character is the ancient graph and the traditional one is a later elaboration, so converting the whole text first would swap an attested form for an unattested one. In the opposite direction, 電 has a seal form while its simplified 电 does not, so the converter resolves 电 through 電 on its own. Neither substitution is applied globally; each character is looked up on its own terms.
That design decision is directly relevant across Southeast Asia. Singapore and Malaysia both have large Chinese-speaking populations, and Chinese calligraphy and seal carving (篆刻) are taught and practised in schools, clan associations and art societies in both countries — the script cut into a stone name chop is exactly this script. But Singapore adopted simplified characters while Taiwan and Hong Kong kept traditional ones, so a Singaporean student setting their own name for a chop and a Taiwanese one are starting from different written forms of the same word. Per-character resolution is what lets both get a correct result without either of them needing to know which graph happens to be the older one.
What you are actually looking at
The glyphs on screen are drawn by 崇羲篆體 (Chong Xi Small Seal), the work of 王心怡 and 季旭昇, published by 中央研究院 (Academia Sinica) and released under a Creative Commons licence that permits commercial use but forbids modification — so the file is served here exactly as they published it. The typeface can draw 11,604 characters, but this converter deliberately draws only 10,387 of them: the ones the 大徐本《說文解字》 actually records. The other 1,217 are forms its designers reconstructed from Shuowen components for characters the dictionary never recorded, which is a legitimate scholarly exercise and the wrong thing for a tool that promises not to invent history. Filtering them out is the single most important thing this converter does. The scholarly layer — headwords, radical assignments and the Shuowen's own glosses — comes from the open shuowen.org dataset of 9,833 entries, released under Apache-2.0.
Treat the output as a typographic rendering, not an epigraphic facsimile. A font makes every stroke uniform; a real seal character cut into stone, or brushed on silk, varies with the hand, the tool and the surface. For learning the shapes, seeing how a radical behaves, or laying out a name before you pick up a knife, this is exactly the right level of detail. For palaeographic research, go to the rubbings and the plates.
10 Things to Know About Small Seal Script
Small seal script was standardised c. 221 BCE under Qin Shi Huang; the chancellor 李斯 (Li Si) is traditionally credited with the standardisation.
The seal forms survive chiefly through the 說文解字, completed by 許慎 (Xu Shen) around 100 CE — a Han dictionary, not a Qin inventory.
This converter draws 10,387 characters — every one recorded in the 說文解字. Its typeface, 崇羲篆體, can draw 1,217 more, and we refuse those: they are modern reconstructions, not recorded forms.
The scholarly layer comes from the open shuowen.org dataset — 9,833 說文解字 entries, released under Apache-2.0.
Simplified characters (简化字) were promulgated from 1956 — over two thousand years later. They were never written in seal script.
Measured on a 400-idiom corpus: 93.5% of characters resolve to an attested seal form, and 77.3% of four-character idioms resolve completely.
The converter refuses to draw a glyph that is not attested. It shows a blank bamboo slip instead — that refusal is the tool's central design claim.
Counter-intuitive: 采 (to pick) has a seal form but its traditional elaboration 採 does not — 采 is the older graph, so a naive 简-to-繁 pass can destroy coverage.
The opposite case: 電 has a seal form and its simplified 电 does not, so the converter resolves 电 through 電 automatically.
艦 (warship), 氧 (oxygen), 硅 (silicon) and 咖啡 (coffee) have no seal form — the things they name post-date the Qin.
Frequently Asked Questions
- Small seal script is the standardised Chinese writing system imposed after the Qin unification around 221 BCE, with the chancellor 李斯 (Li Si) traditionally credited for reducing the competing regional scripts to one. Its rounded, evenly balanced, vertically composed forms are the ancestors of every later Chinese script. It is still the script used for carved name seals and chops today, which is why calligraphers and seal carvers keep learning it.
- No, and we are careful never to claim that. What you see is the form recorded in the 說文解字, the dictionary completed by 許慎 (Xu Shen) around 100 CE — a Han-dynasty scholar's record of the Qin standard, written three centuries after it was set. The Shuowen also carries the 新附 characters appended by 徐鉉 (Xu Xuan) in 986 CE, so it is a dictionary with its own long history rather than a Qin inventory. The correct description of every glyph here is "the form recorded in the 說文解字".
- There are two different reasons. The first is that the character is a modern coinage or a simplified form that was simply never written in seal script — 氧 (oxygen) and 咖啡 (coffee) name things the Qin did not have, and simplified characters date from 1956. The second is subtler: the 說文解字 sometimes records the word under an older headword spelling that our typeface indexes as a separate character, so an attested form exists in principle but not under the character you typed. Our sources can also simply be incomplete, and a blank is how we say so rather than guessing.
- Copy gives you the modern Chinese text — traditional or simplified as resolved — not seal glyphs. The seal appearance here is a typeface rendering ordinary Chinese characters, not a separate set of characters with their own codepoints, so there is nothing "seal" to put on the clipboard. Paste it anywhere and it will display in whatever font the destination uses, which will almost always be a normal modern one. If you want the seal look to travel with you, download the image instead.
- It is a good starting point for laying out a name and seeing which characters have an attested seal form before you commit to stone. What it will not do is make the compositional decisions a carver makes — spacing, stroke weight, how characters are squeezed or stretched to fill a square, and the choice between 朱文 and 白文. Treat the output as a reference sheet, and take it to someone who carves if the seal matters.
- Yes, and it does so honestly. Simplified characters were promulgated from 1956, so as shapes they have no seal form of their own, but many of them correspond to a traditional character that does — 电 is resolved through 電, which is attested. Where no such route exists the tool shows a blank rather than a synthesised glyph. So simplified input works well, but it is worth knowing that what you are seeing is the seal form of the corresponding older character.
- No — that can make things worse. It is tempting to assume traditional characters are always the older and better-attested ones, but 采 (to pick) has a seal form while its traditional elaboration 採 does not, because 采 is the older graph. A blanket 简-to-繁 pass would swap an attested form for an unattested one and destroy coverage. This converter resolves each character independently in both directions, so mixed text is fine and you should paste it exactly as you have it.
- We measured it rather than estimating it. On a corpus of 400 four-character idioms, 93.5% of individual characters resolved to an attested seal form and 77.3% of the idioms resolved completely, meaning all four characters were attested. The gate itself admits 10,387 characters — the ones the 大徐本《說文解字》 records. Everyday classical and literary text does well; modern technical, scientific and loanword vocabulary does not, and that gap is real rather than a bug.
- The seal glyphs are drawn with 崇羲篆體 (Chong Xi Small Seal) by 王心怡 and 季旭昇, published by 中央研究院 (Academia Sinica) under CC BY-ND 3.0 TW or later. The typeface covers 11,604 characters, but this converter only draws the 10,387 the 大徐本《說文解字》 actually records — the other 1,217 are modern transcriptions built from Shuowen components, and drawing them would be inventing history. The headwords, radical assignments and Shuowen glosses come from the open shuowen.org dataset of 9,833 說文解字 entries, released under Apache-2.0. All are credited in full at the foot of this page.
- It is a typographic rendering, not an epigraphic facsimile: a font makes every stroke uniform, while a real seal character cut in stone or brushed on silk varies with the hand, the tool and the surface. For learning, teaching, and laying out a design it is exactly right. For palaeographic research consult rubbings, plates and the critical editions, and for anything permanent get a reading from someone who works in the script — a character that is correct is not always the character you meant.
Related News
You may be interested in these recent stories from our newsroom.
No related news yet for this tool. Our editorial team publishes new pieces every week.
Browse all news →Pick up where you left off
Stored only in this browser — never sent to our servers.