Artha v1 E2B (LiteRT, on-device)
Run Artha v1 E2B fully offline on your phone. Type a rough request in English, Hinglish or Gujlish and get back a ready-to-send email, WhatsApp message or LinkedIn post that reads like a real person wrote it.
After the one-time download, nothing leaves your device.
Install on your phone
- Install Google AI Edge Gallery from the Play Store or the App Store. Requires Android 12+ or iOS 17+.
- In the model list, choose import from Hugging Face.
- Enter
developerJenis/Artha-v1-E2B-LiteRT. - Choose the GPU backend for the fastest replies.
Recommended: a phone with 6 GB of RAM or more. The model file is about 2.8 GB (4-bit weights).
Try these
client ne 45 din se 2.4 lakh ka payment nahi kiya, firm mail likho. client Mehta Traders, contact Rakesh, mera naam Jenis
friend ko bolna hai aaj party me nahi aa paunga, office me late ho gaya
leave mail to my manager Sunil sir, 12 to 14 Dec, sister's wedding in Rajkot. I'm Priya
Include the real details you want used (names, amounts, dates) and check them before sending.
Desktop
The same .litertlm file runs on Linux, macOS and Windows with the LiteRT-LM CLI:
pip install litert-lm
litert-lm run --from-huggingface-repo=developerJenis/Artha-v1-E2B-LiteRT Artha-v1-E2B.litertlm \
--prompt="friend ko bolna hai aaj party me nahi aa paunga, office me late ho gaya"
For Ollama and llama.cpp use Artha-v1-E2B-GGUF. Full details, evaluation and limitations are on the main model card.
License
Apache 2.0, the same license as the Gemma 4 base model. Built by Jenis (developerJenis).
- Downloads last month
- 10