01
Best at
Mien (Latin script) into English. Most accurate direction.
01
eng_Latn0 chars
02
ium_Latn0 chars
English ↔ Iu Mien (Latin), strongest on historical baseline text.
Mien (Latin script) into English. Most accurate direction.
Uses the unified Latin orthography (1984), the script most Iu Mien speakers read and write today.
Modern slang, jokes, and very recent vocabulary may come out off.
The v7 model scores chrF 62 / BLEU 47 on a held-out test set for ium → eng. The reverse direction (eng → ium) lands at chrF 29 / BLEU 12 — harder for any low-resource MT model. Best on the unified 1984 Latin orthography and on everyday vocabulary; weakest on slang, regional dialects, and very long passages.
English ↔ Iu Mien (Latin script, ium_Latn). The model was fine-tuned from Meta's NLLB-200-distilled-1.3B on 179,386 parallel sentence pairs covering both directions. Latin script is what most modern Iu Mien speakers read and write today.
Yes. The translator at learniumien.org/translator is free and open to anyone without an account. The underlying API at api.learniumien.org/translate accepts JSON requests with no authentication; rate limits apply but no key required.
The modern Iu Mien standard is the Unified Mien Romanization (often called IuMiNR), adopted in 1984 — a Latin alphabet with tone letters (-c, -h, -v, -x, -z) at word ends. Older texts also use Han characters, Thai script, and historical Lao script. The translator targets the modern Latin orthography.