Built-in languages and default models

Use this page to choose a language alias, Java enum, default model ID, and aggregate package. The exhaustive artifact provenance and checksums live in the generated model catalog; executable loading examples live in Model Selection and Loading.

“Registered language” means that the repository maintains a dictionary and a stable Java runtime mapping for that language. Packaging differs by runtime: Java keeps the core dictionary-free and resolves individual model artifacts, while the Java and Python radixor-models-standard aggregates install the same reviewed 31 standard dictionaries.

The Java language enum carries language identity, writing direction, a legacy resource-directory name, and the stable default model ID. A Java model descriptor carries the independently versioned model identity and resource. Python accepts the short alias or the same full model ID. See Model Selection and Loading for Java and Python Usage and API for Python.

Registered languages and models

The Java and Python standard aggregates contain the 31 IDs in models/standard-model-projects.properties. Every other active model is individually published and belongs to the Java-only extended aggregate; filtered alternatives use their separate opt-in Java aggregate.

Language	Java enum	Model ID	Topology role	Package	Relative dictionary size
Adyghe	ADY	ady-default	standalone	extended	★★★★☆ (20,347 distinct forms)
Afrikaans	AF_ZA	af-za-default	standalone	extended	★★★★★ (297,154 distinct forms)
Aimele	AIL	ail-default	standalone	extended	★★☆☆☆ (3,061 distinct forms)
Akan	AK	aka-default	standalone	extended	★★☆☆☆ (1,890 distinct forms)
Albanian	SQ_AL	sq-al-default	standalone	extended	★★★☆☆ (10,218 distinct forms)
Alsatian	GSW	gsw-default	standalone	extended	★★☆☆☆ (1,186 distinct forms)
Amharic	AM_ET	am-et-default	standalone	extended	★★★★☆ (41,308 distinct forms)
Amuzgo	AZG	azg-default	standalone	extended	★★★☆☆ (9,371 distinct forms)
Ancient Greek	GRC	grc-default	standalone	extended	★★★★☆ (26,386 distinct forms)
Anglo-Norman	XNO	xno-default	standalone	extended	★☆☆☆☆ (186 distinct forms)
Arabic	AR	ar-default	standalone	standard	★★★★★ (452,974 distinct forms)
Armenian	HY_AM	hy-am-default	standalone	standard	★★★★★ (246,576 distinct forms)
Ashaninka	CNI	cni-default	standalone	extended	★★★☆☆ (10,095 distinct forms)
Assamese	AS_IN	as-in-default	standalone	extended	★★★★★ (67,733 distinct forms)
Asturian	AST	ast-default	standalone	extended	★★★★☆ (21,589 distinct forms)
Aymara	AYM	aym-default	standalone	extended	★★★★★ (308,884 distinct forms)
Azerbaijani	AZ_AZ	az-az-default	standalone	extended	★★★☆☆ (6,669 distinct forms)
Bashkir	BAK	bak-default	standalone	extended	★★★☆☆ (10,723 distinct forms)
Belarusian	BE_BY	be-by-default	standalone	extended	★★★★☆ (19,527 distinct forms)
Bengali	BN_BD	bn-bd-default	standalone	extended	★★☆☆☆ (2,847 distinct forms)
Bininj Kun-wok	GUP	gup-default	standalone	extended	★☆☆☆☆ (370 distinct forms)
Braj	BRA	bra-default	standalone	extended	★★☆☆☆ (2,028 distinct forms)
Breton	BRE	bre-default	standalone	extended	★★☆☆☆ (1,940 distinct forms)
Bulgarian	BG_BG	bg-bg-default	standalone	extended	★★★★☆ (45,806 distinct forms)
Catalan	CA_ES	ca-es-default	standalone	standard	★★★★★ (130,366 distinct forms)
Cebuano	CEB	ceb-default	standalone	extended	★☆☆☆☆ (430 distinct forms)
Chichewa	NY_MW	ny-mw-default	standalone	extended	★★★☆☆ (3,178 distinct forms)
Chichicapan Zapotec	ZPV	zpv-default	standalone	extended	★★☆☆☆ (818 distinct forms)
Chukchi	CKT	ckt-default	standalone	extended	★☆☆☆☆ (302 distinct forms)
Church Slavonic	CHU	chu-default	standalone	extended	★★☆☆☆ (1,652 distinct forms)
Classical Armenian	XCL	xcl-default	standalone	extended	★★★★☆ (59,215 distinct forms)
Classical Syriac	SYC	syc-default	standalone	extended	★★★★☆ (27,529 distinct forms)
Congo Swahili	SWC	swc-default	standalone	extended	★★★☆☆ (11,121 distinct forms)
Copala Triqui	CPA	cpa-default	standalone	extended	★★☆☆☆ (2,741 distinct forms)
Cornish	COR	cor-default	standalone	extended	★☆☆☆☆ (162 distinct forms)
Cree	CRE	cre-default	standalone	extended	★★☆☆☆ (1,076 distinct forms)
Crimean Tatar	CRH	crh-default	standalone	extended	★★★☆☆ (7,195 distinct forms)
Czech	CS_CZ	cs-cz-default	default	standard	★★★★☆ (51,401 distinct forms)
Dakota	DAK	dak-default	standalone	extended	★★★☆☆ (3,332 distinct forms)
Danish	DA_DK	da-dk-default	default	standard	★★★★☆ (27,921 distinct forms)
Dutch	NL_NL	nl-nl-default	default	standard	★★★★☆ (26,201 distinct forms)
Eastern Chatino	CLY	cly-default	standalone	extended	★★☆☆☆ (2,035 distinct forms)
Egyptian Arabic	ARZ	arz-default	standalone	extended	★★★★☆ (17,937 distinct forms)
English	US_UK	us-uk-default	default	standard	★★★★★ (591,946 distinct forms)
Estonian	ET_EE	et-ee-default	standalone	standard	★★★★☆ (24,811 distinct forms)
Evenki	EVN	evn-default	standalone	extended	★★★☆☆ (13,549 distinct forms)
Faroese	FO_FO	fo-fo-default	standalone	extended	★★★★☆ (31,366 distinct forms)
Finnish	FI_FI	fi-fi-default	default	standard	★★★★★ (1,788,784 distinct forms)
French	FR_FR	fr-fr-default	default	standard	★★★★★ (404,011 distinct forms)
Friulian	FUR	fur-default	standalone	extended	★★★☆☆ (5,007 distinct forms)
Ga	GAA	gaa-default	standalone	extended	★☆☆☆☆ (469 distinct forms)
Galolen	GAL	gal-default	standalone	extended	★★★★☆ (25,436 distinct forms)
German	DE_DE	de-de-default	default	standard	★★★★★ (277,266 distinct forms)
Gothic	GOT	got-default	standalone	extended	★★★★★ (134,332 distinct forms)
Greek	EL_GR	el-gr-default	standalone	standard	★★★★★ (76,869 distinct forms)
Gulf Arabic	AFB	afb-default	standalone	extended	★★★★☆ (29,344 distinct forms)
Haida	HAI	hai-default	standalone	extended	★★★☆☆ (5,385 distinct forms)
Hebrew	HE_IL	he-il-default	default	standard	★★★★☆ (57,658 distinct forms)
Hiligaynon	HIL	hil-default	standalone	extended	★☆☆☆☆ (393 distinct forms)
Hsilimo	HSI	hsi-default	standalone	extended	★☆☆☆☆ (158 distinct forms)
Hungarian	HU_HU	hu-hu-default	default	standard	★★★★★ (910,688 distinct forms)
Icelandic	IS_IS	is-is-default	standalone	extended	★★★★☆ (52,197 distinct forms)
Indonesian	ID_ID	id-id-default	standalone	standard	★★★★☆ (21,296 distinct forms)
Ingrian	IZH	izh-default	standalone	extended	★★☆☆☆ (1,026 distinct forms)
Irish	GA_IE	ga-ie-default	standalone	standard	★★★★☆ (24,035 distinct forms)
Italian	IT_IT	it-it-default	default	standard	★★★★★ (324,366 distinct forms)
Itelmen	ITL	itl-default	standalone	extended	★★★☆☆ (3,659 distinct forms)
Japanese	JA_JP	ja-jp-default	standalone	extended	★★★☆☆ (10,848 distinct forms)
Kabardian	KBD	kbd-default	standalone	extended	★★☆☆☆ (3,054 distinct forms)
Kalaallisut	KL_GL	kl-gl-default	standalone	extended	★☆☆☆☆ (321 distinct forms)
Kannada	KN_IN	kn-in-default	standalone	extended	★★★☆☆ (3,802 distinct forms)
Karelian	KRL	krl-default	standalone	extended	★☆☆☆☆ (566 distinct forms)
Kashubian	CSB	csb-default	standalone	extended	★☆☆☆☆ (350 distinct forms)
Kazakh	KK_KZ	kk-kz-default	standalone	extended	★★★★☆ (35,223 distinct forms)
Khakas	KJH	kjh-default	standalone	extended	★★☆☆☆ (1,172 distinct forms)
Khaling	KLR	klr-default	standalone	extended	★★★★☆ (54,602 distinct forms)
Kodi	KOD	kod-default	standalone	extended	★☆☆☆☆ (524 distinct forms)
Kongo	KON	kon-default	standalone	extended	★☆☆☆☆ (557 distinct forms)
Kyrgyz	KY_KG	ky-kg-default	standalone	extended	★★☆☆☆ (2,997 distinct forms)
Ladin	LLD	lld-default	standalone	extended	★★★☆☆ (4,819 distinct forms)
Latin	LA	la-default	standalone	extended	★★★★★ (486,375 distinct forms)
Latvian	LV_LV	lv-lv-default	standalone	extended	★★★★★ (75,492 distinct forms)
Lingala	LIN	lin-default	standalone	extended	★☆☆☆☆ (230 distinct forms)
Lithuanian	LT_LT	lt-lt-default	standalone	standard	★★★★☆ (28,889 distinct forms)
Livonian	LIV	liv-default	standalone	extended	★★☆☆☆ (2,861 distinct forms)
Low German	NDS	nds-default	standalone	extended	★★☆☆☆ (2,495 distinct forms)
Lower Sorbian	DSB	dsb-default	standalone	extended	★★★☆☆ (12,039 distinct forms)
Luganda	LG_UG	lg-ug-default	standalone	extended	★★★☆☆ (4,673 distinct forms)
Macedonian	MK_MK	mk-mk-default	standalone	extended	★★★★★ (135,754 distinct forms)
Magahi	MAG	mag-default	standalone	extended	★★☆☆☆ (1,448 distinct forms)
Malagasy	MG_MG	mg-mg-default	standalone	extended	★☆☆☆☆ (636 distinct forms)
Maltese	MT_MT	mt-mt-default	standalone	extended	★★☆☆☆ (1,491 distinct forms)
Manx	GV_IM	gv-im-default	standalone	extended	★☆☆☆☆ (15 distinct forms)
Maori	MI_NZ	mi-nz-default	standalone	extended	★☆☆☆☆ (207 distinct forms)
Mapudungun	ARN	arn-default	standalone	extended	★☆☆☆☆ (548 distinct forms)
Middle French	FRM	frm-default	standalone	extended	★★★★☆ (27,102 distinct forms)
Middle High German	GMH	gmh-default	standalone	extended	★☆☆☆☆ (384 distinct forms)
Middle Low German	GML	gml-default	standalone	extended	★☆☆☆☆ (624 distinct forms)
Mongolian	MN_MN	mn-mn-default	standalone	extended	★★★★☆ (17,231 distinct forms)
Murrinh-Patha	MWF	mwf-default	standalone	extended	★☆☆☆☆ (592 distinct forms)
Navajo	NAV	nav-default	standalone	extended	★★★☆☆ (11,046 distinct forms)
Neapolitan	NAP	nap-default	standalone	extended	★★☆☆☆ (1,497 distinct forms)
North Frisian	FRR	frr-default	standalone	extended	★☆☆☆☆ (364 distinct forms)
Northern Sami	SME	sme-default	standalone	extended	★★★★☆ (54,435 distinct forms)
Norwegian Bokmål	NB_NO	nb-no-default	default	standard	★★★★★ (73,170 distinct forms)
Norwegian Nynorsk	NN_NO	nn-no-default	default	standard	★★★★☆ (16,937 distinct forms)
Old English	ANG	ang-default	standalone	extended	★★★★☆ (65,202 distinct forms)
Old French	FRO	fro-default	standalone	extended	★★★★★ (94,338 distinct forms)
Old High German	GOH	goh-default	standalone	extended	★★★☆☆ (5,267 distinct forms)
Old Irish	SGA	sga-default	standalone	extended	★★☆☆☆ (934 distinct forms)
Old Norse	NON	non-default	standalone	extended	★★★★☆ (46,089 distinct forms)
Old Saxon	OSX	osx-default	standalone	extended	★★★☆☆ (12,361 distinct forms)
Oodham	OOD	ood-default	standalone	extended	★★☆☆☆ (1,159 distinct forms)
Otomi (ote)	OTE	ote-default	standalone	extended	★★★☆☆ (3,417 distinct forms)
Pashto	PS_AF	ps-af-default	standalone	extended	★★☆☆☆ (2,945 distinct forms)
Persian	FA_IR	fa-ir-default	default	standard	★★★☆☆ (3,544 distinct forms)
Polish	PL_PL	pl-pl-polimorf	optional	extended	★★★★★ (4,668,685 distinct forms)
Polish	PL_PL	pl-pl-unimorph	default	standard	★★★★★ (120,867 distinct forms)
Portuguese	PT_PT	pt-pt-default	default	standard	★★★★★ (211,091 distinct forms)
Quechua	QUE	que-default	standalone	extended	★★★★★ (147,914 distinct forms)
Romanian	RO_RO	ro-ro-default	standalone	standard	★★★★☆ (48,565 distinct forms)
Russian	RU_RU	ru-ru-default	default	standard	★★★★★ (759,333 distinct forms)
Seneca	SEE	see-default	standalone	extended	★★★☆☆ (5,318 distinct forms)
Serbo-Croatian	HBS	hbs-default	standalone	extended	★★★★★ (272,634 distinct forms)
Shipibo-Conibo	SHP	shp-default	standalone	extended	★★★☆☆ (7,729 distinct forms)
Shona	SN_ZW	sn-zw-default	standalone	extended	★★☆☆☆ (2,640 distinct forms)
Sotho, Southern	ST_ZA	st-za-default	standalone	standard	★☆☆☆☆ (416 distinct forms)
Southern Kurdish	SDH	sdh-default	standalone	extended	★☆☆☆☆ (165 distinct forms)
Spanish	ES_ES	es-es-default	default	standard	★★★★★ (849,661 distinct forms)
Swedish	SV_SE	sv-se-default	default	standard	★★★★★ (95,181 distinct forms)
Tagalog	TL_PH	tl-ph-default	standalone	extended	★★☆☆☆ (2,510 distinct forms)
Turkish	TR_TR	tr-tr-default	standalone	standard	★★★★★ (222,553 distinct forms)
Ukrainian	UK_UA	uk-ua-default	default	standard	★★★★☆ (14,150 distinct forms)
Uyghur	UG_CN	ug-cn-default	standalone	extended	★★★☆☆ (6,368 distinct forms)
Uzbek	UZ_UZ	uz-uz-default	standalone	extended	★★★☆☆ (9,835 distinct forms)
Voro	VRO	vro-default	standalone	extended	★☆☆☆☆ (412 distinct forms)
Western Highland Chatino	CTP	ctp-default	standalone	extended	★★☆☆☆ (2,597 distinct forms)
Xibe	SJO	sjo-default	standalone	extended	★★★☆☆ (3,151 distinct forms)
Yanesha’	AME	ame-default	standalone	extended	★★☆☆☆ (2,635 distinct forms)
Yiddish	YI	yi-default	default	standard	★★★☆☆ (3,532 distinct forms)
Yoloxóchitl Mixtec	XTY	xty-default	standalone	extended	★★☆☆☆ (2,754 distinct forms)
Zarma	DJE	dje-default	standalone	extended	★☆☆☆☆ (81 distinct forms)
Zenzontepec Chatino	CZN	czn-default	standalone	extended	★★☆☆☆ (1,864 distinct forms)
Zulu	ZU_ZA	zu-za-default	standalone	extended	★★★★☆ (32,384 distinct forms)

The maintained table deliberately avoids duplicating mutable provenance fields. Those values come from model module metadata and the generated model catalog.

The Polish dual-model case

PL_PL represents Polish. It is not an alias for either source dictionary.

• loadCompiled(Language.PL_PL, ...) resolves pl-pl-unimorph.
• registry.require("pl-pl-polimorf") resolves the optional PoliMorf model.
• StemmerPatchTrieLoader.loadCompiled("pl-pl-polimorf", true, reductionMode) constructs its compiled trie explicitly; complete construction is verified with a dedicated 6 GiB test heap.
• Both artifacts may be present and loaded independently.
• Adding PoliMorf does not change the language default.
• Radixor does not merge their dictionaries or outputs automatically.

UniMorph and PoliMorf have different lexical sources and provenance. Applications should compare outputs with application-specific regression tests before changing an explicit model choice.

In Python, Stemmer("pl") selects pl-pl-unimorph. The standard Python data package does not include PoliMorf; applications that need it must compile and load it explicitly as a trusted custom model. As in Java, it never changes the Polish default implicitly.

Dependency patterns

Resolve the placeholders below from the Maven Central artifact page and the model catalog, respectively.

Minimal English:

dependencies {
    implementation 'org.egothor:radixor:<latest-java-version>'
    runtimeOnly 'org.egothor:radixor-model-us-uk-default:<compatible-model-version>'
}

All 31 standard-aggregate defaults:

dependencies {
    implementation 'org.egothor:radixor:<latest-java-version>'
    runtimeOnly 'org.egothor:radixor-models-standard:<compatible-catalog-version>'
}

The standard pack is metadata-only and excludes optional PoliMorf.

Every individual model artifact carries its own provenance and licensing material. UniMorph models carry model-specific audited licenses and notices because their official language repositories identify different lexical sources, contributors, and applicable terms. Each notice preserves upstream attribution and records the Radixor transformations and Leo Galambos contribution statement. Legacy imports disclose when an exact historical revision was not recorded; this is a reproducibility limitation, not a claim that the source or license is unknown. The active new-model set includes CC BY-SA 3.0, CC BY-SA 4.0, CC BY 4.0, and LGPLLR material; unsupported evidence is quarantined rather than assigned a generic license.

Loading a language default

final FrequencyTrie<CompiledPatchCommand> trie =
        StemmerPatchTrieLoader.loadCompiled(
                StemmerPatchTrieLoader.Language.US_UK,
                true,
                ReductionMode.MERGE_SUBTREES_WITH_EQUIVALENT_RANKED_GET_ALL_RESULTS);

The call discovers the default descriptor from the runtime classpath, verifies its compressed resource, parses the GZip UTF-8 dictionary, and constructs a read-only trie. A missing default throws StemmerModelNotFoundException; there is no arbitrary fallback.

Writing direction

Arabic (including Gulf and Egyptian Arabic), Persian, Hebrew, Pashto, Southern Kurdish, Classical Syriac, Uyghur, Urdu, and Yiddish declare right-to-left writing metadata for presentation. Writing direction does not reorder characters in a Java String: all built-in natural-language models therefore use backward traversal from the stored sequence end, where suffixes remain located. Explicit forward traversal is reserved for deliberately prefix-oriented custom data. The selected traversal must remain aligned across dictionary parsing, trie lookup, patch generation, persistence, and application; model identity and writing direction are separate concerns.

Custom and persisted alternatives

Registered model artifacts are a convenient reproducible baseline. Applications may instead load caller-owned textual dictionaries or persist compiled .radixor.gz tries. Those paths are distinct from model artifact discovery:

• a model stemmer.gz is a compressed textual dictionary plus descriptor/index metadata;
• a .radixor.gz created by the binary writer is a persisted compiled trie;
• a source dictionary is upstream input, not automatically a valid model artifact.

See Dictionary Format, CLI Compilation, and Stemmer Models.

Benchmark interpretation

Benchmark rows must identify the Radixor model ID used. Default rows use the default IDs above. Optional Polish PoliMorf comparisons must be labeled pl-pl-polimorf; they are not interchangeable with the historical default Polish row. Continue with Benchmarking and Reproducibility.