Language Support
Playto supports 12 UI languages, but feature quality varies per language depending on the underlying open-source models. This page lists the current support tier for each feature, plus a reference of widely-used proficiency exams.
Last updated: May 23, 2026
Tier definitions
Feature matrix
Status as of the current release. Tiers will be upgraded as upcoming TTS overhaul work for CJK languages ships.
| Language | OCR | Translation | TTS | Pronunciation | Dictionary |
|---|---|---|---|---|---|
| English | Stable | Stable | Stable | Stable | Stable |
| 日本語 1 | Stable | Stable | Functional | Beta | Stable |
| 中文 (简) 2 | Stable | Stable | Unsupported | Beta | Stable |
| 한국어 2 | Stable | Stable | Unsupported | Beta | Stable |
| Español 3 | Stable | Stable | Functional | Beta | Stable |
| Français 3 | Stable | Stable | Functional | Beta | Stable |
| Italiano 3 | Stable | Stable | Functional | Beta | Stable |
| Português (BR) 3 | Stable | Stable | Functional | Beta | Stable |
| Deutsch 2 | Stable | Stable | Unsupported | Beta | Stable |
| Русский 2 | Stable | Stable | Unsupported | Beta | Stable |
| Polski 2 | Stable | Stable | Unsupported | Beta | Stable |
| Tiếng Việt 2 | Stable | Stable | Unsupported | Beta | Stable |
- Dedicated voice available; pitch accent expression has limits — improvement planned via the upcoming TTS overhaul.
- The current TTS engine does not support this language, so the in-app TTS button is disabled. Integration of a dedicated TTS engine for this language is planned for a future update.
- TTS voice mapping was recently extended. Not native quality, but practical for learning use.
Dictionary backing
Word definitions shown in Playto are currently generated by the LLM at lookup time across all 12 languages. This works well in practice but is not backed by an authoritative dictionary. External dictionary integration (JMdict for Japanese, CC-CEDICT for Chinese, ECDICT-style sources for English and others) is on the roadmap to reduce hallucination and improve definition consistency.
Idiom and phrasal expression coverage
Idiom and multi-word expression detection is not yet implemented in any language. Single-word lookup works across all 12 languages; multi-word idioms may not be recognized as units. Idiom dictionary integration per language is a roadmap candidate.
Lightweight NMT translation
Alongside the AI translation modes, Playto includes Lightweight NMT — a compact neural machine translation engine. It translates in tens of milliseconds and runs on minimal hardware, making it a fast, low-VRAM option. Its quality ceiling is lower than the local AI models, and it covers a fixed set of language pairs. For any pair not listed below, Playto uses its AI translation instead.
| Direction | NMT support |
|---|---|
| Japanese ↔ English | Both directions |
| English → Spanish / French / Italian / Portuguese / German / Polish | One direction (English as source) |
| All other language pairs | AI translation only |
Proficiency exam reference
Major standardized exams per language. Japanese, English, and Chinese vocabulary is currently tagged with JLPT / CEFR / HSK level data internally. Pack-level estimation and level-aware SRS difficulty are planned for a future update across all supported languages.
| Language | Major exams | Levels | Playto today | Future |
|---|---|---|---|---|
| English | CEFR (A1–C2), TOEIC, TOEFL, IELTS | 6 (CEFR) | Vocabulary tagged (KELLY EVP, 6,481 entries) | Pack-level mapping, level-aware SRS |
| 日本語 | JLPT (N5–N1), J.TEST | 5 (JLPT) | Vocabulary tagged (JLPT N5–N1, 32,307 entries) | Pack-level mapping, level-aware SRS |
| 中文 (简) | HSK 3.0 (1–9) | 9 | Vocabulary tagged (HSK 3.0, 5,342 entries) | Pack-level mapping, level-aware SRS |
| 한국어 | TOPIK (1–6) | 6 | Not integrated | TOPIK vocabulary integration |
| Español | DELE / SIELE (CEFR A1–C2) | 6 | Not integrated | CEFR vocabulary integration |
| Français | DELF / DALF (CEFR A1–C2), TCF | 6 | Not integrated | CEFR vocabulary integration |
| Italiano | CILS, CELI (CEFR A1–C2) | 6 | Not integrated | CEFR vocabulary integration |
| Português (BR) | CELPE-Bras, CIPLE | 4–5 | Not integrated | CEFR-aligned vocabulary integration |
| Deutsch | Goethe-Zertifikat (A1–C2), TestDaF | 6 | Not integrated | CEFR vocabulary integration |
| Русский | TORFL / TRKI (A1–C2) | 6 | Not integrated | TORFL vocabulary integration |
| Polski | Certyfikat (A1–C2) | 6 | Not integrated | CEFR vocabulary integration |
| Tiếng Việt | CEFR-V framework | 6 (CEFR-V) | Not integrated | CEFR-V vocabulary integration |
Roadmap
- Upcoming — CJK TTS overhaul: improved pitch accent for Japanese, tone marker integration for Chinese, and a dedicated TTS engine for Korean.
- Future — DE / RU / PL / VI TTS: dedicated TTS engine integration based on user demand.
- Future — Proficiency level estimation: per-pack level metadata (CEFR / JLPT / HSK / TOPIK), level-aware SRS difficulty.
- Future — Proficiency vocabulary expansion: TOPIK (Korean), CEFR (DE / ES / FR / IT / PT / PL), TORFL (Russian), CEFR-V (Vietnamese) vocabulary integration to bring all 12 languages to parity with JA / EN / ZH.
- Future — External dictionary integration: JMdict (JA), CC-CEDICT (ZH), ECDICT-style sources (EN and others) to back word definitions with authoritative data instead of LLM-only generation.
- Future — Idiom and phrasal expression detection: per-language idiom dictionaries for multi-word expression recognition.
Found an inaccuracy? Open an issue.