🪴 hachiue
Mobile WebGPU LLM & Translation
GitHub
Bonsai 1.7B Q1_0 (237 MiB)
Bonsai 4B Q1_0 (546 MiB)
Bonsai 8B Q1_0 (1,105 MiB)
Bonsai 2 2B PTQ1_0 (409 MiB)
Bonsai 2 2B PQ2_0 (493 MiB)
Bonsai 2 4B PTQ1_0 (906 MiB)
Bonsai 2 4B PQ2_0 (1,093 MiB)
Ternary Bonsai 1.7B PQ2_0 (442 MiB)
Ternary Bonsai 4B PQ2_0 (1,025 MiB)
PLaMo 3 Translate 0.8B Q4_0 (522 MiB)
PLaMo 3 Translate 0.8B Q4_K_M (546 MiB)
PLaMo 3 Translate 2B Q4_0 (1,620 MiB)
PLaMo 3 Translate 2B Q4_K_M (1,723 MiB)
Load Model
Prefetch
Reset Cache
Select a PLaMo 3 Translate or Bonsai model and tap "Load Model".
Remove downloaded models
💬 Chat
🌐 Translate (Text & Term)
📄 Document / Sentence (translateText)
📖 Word / Phrase Lookup (translateTerm)
From
English (en)
Japanese (ja)
Chinese Simplified (zh-CN)
Chinese Traditional (zh-TW)
Korean (ko)
Spanish (es)
French (fr)
German (de)
→
To
Japanese (日本語)
Easy Japanese (やさしい日本語)
English
Chinese Simplified (简体中文)
Chinese Traditional (繁體中文)
Korean (한국어)
Spanish (Español)
French (Français)
German (Deutsch)
Style: Auto
Polite (です・ます)
Plain (だ・である)
Source Text
Small trees thrive in small pots. With 1-bit and 4-bit quantization on WebGPU, language models run directly inside mobile browsers.
Target Word or Short Phrase
Optional Context (surrounding sentence or domain note)
Optional Glossary (`source=target`, comma-separated)
Translate
Translation Output