The Open Weights
LatestModelsLeaderboardsCompanies
Subscribe
The Open Weights

The daily record of open-source AI. New model releases, leaderboards, and what's coming next — written for people who ship.

Refreshed every 12 hours

Discover

  • Latest releases
  • New today
  • Trending models

Browse

  • All models
  • Companies
  • Categories
  • Leaderboards

About

  • About
  • Editorial policy
  • Hardware estimates
  • RSS feed
  • llms.txt
  • Newsletter

© 2026 The Open Weights. An independent publication.

PrivacyTermsSMSAggregated by Claude · curated by humans.
LatestOpenMOSS1.0
OpenMOSSText / LLM

MameLoshnLM Brings Yiddish Into the Open LLM Era

OpenMOSS releases what it calls the first open-source 8B language model built specifically for Yiddish, paired with an evaluation benchmark.

Aug 5, 2026
UpdateOther

OpenMOSS has released MameLoshnLM, an 8-billion-parameter language model built specifically for Yiddish, along with a benchmark for measuring how well systems handle the language. According to the accompanying paper on Hugging Face, it is positioned as the first open-source model tailored to Yiddish rather than treating it as a rare afterthought in a larger multilingual mix.

The name itself is a nod to the language it serves: mame-loshn — literally "mother tongue" — is the affectionate Yiddish term for Yiddish. That framing matters, because Yiddish is a low-resource language for machine learning purposes, with far less digitized text than English, Mandarin, or the major European languages that dominate today's training corpora.

Why it matters

Most general-purpose large language models absorb dozens of languages, but coverage thins out quickly for smaller communities. A dedicated model and benchmark give researchers and speakers something they have largely lacked:

  • An open 8B model sized to run on modest hardware, not just data-center clusters.
  • A shared evaluation benchmark so future Yiddish systems can be compared consistently.
  • A reference point for applying similar work to other under-served languages.

The release lands in the mid-sized 7B–13B tier that has become the practical sweet spot for open models, balancing capability against the cost of fine-tuning and deployment. The license is listed as "other," so teams considering commercial or derivative use should review the terms before building on it. Full technical details, including training data and benchmark results, are available in the paper.

Sources

  • MameLoshnLM: Yiddish Language Model and Evaluation Benchmark

    HF Papers

    Visit
OlderQwen3.8-27B Brings Vision to a Dense ModelQwen · Alibaba · Vision-Language · 4 months agoNewerMotif 3 brings a sparse MoE approach to reasoningMotif Technologies · Text / LLM · 3 months ago

Get the model

HF Papers

Specs

Parameters8B
LicenseOTHER

Can you run it?

Runs on a laptop — about 6.9 GB at 4-bit.

  • BF16 (as published)

    24 GB GPU (RTX 3090 / 4090) · 32 GB Mac

    18.1 GB
  • 8-bit

    12 GB GPU (RTX 3060 / 4070) · 16 GB Mac

    10.6 GB
  • 4-bit

    8 GB GPU (RTX 4060) · 16 GB laptop or Mac

    6.9 GB
Your machine
  • Precision not published; assuming 16-bit (BF16) weights.
  • Parameter count is an editorial estimate — no weights metadata on Hugging Face.
  • KV cache approximated from model size (no config metadata) at 8,192 tokens.

Estimated from the parameter count. Assumes an 8K context and ~1 GB runtime overhead; actual needs vary. How we estimate


Modalities

Text / LLM

The Weekly Weights

Every open release that mattered, one email a week.

0 comments

No comments yet. Be the first to weigh in.

More from OpenMOSS

All OpenMOSS releases →
MOSS-Image SFT
OpenMOSS/Text → ImageRuns on a laptop

OpenMOSS Releases MOSS-Image Under Apache 2.0

The academic team behind the MOSS project moves into text-to-image generation with a permissively licensed model and a fine-tuned SFT variant.

Oct 11, 2026
MOSS-Image SFT
OpenMOSS/Text → ImageRuns on a laptop

OpenMOSS Releases Apache-2.0 Bilingual Image Model

MOSS-Image Base arrives as a permissively licensed text-to-image foundation with English and Chinese support.

Oct 11, 2026
OpenMOSS/Vision-Language

OpenMOSS Debuts MOSS-VL for Real-Time Vision Interaction

A new open vision-language model family uses gated cross-attention to enable streaming, low-latency multimodal exchanges.

Aug 14, 2026

More in Text / LLM

All Text / LLM →
Aion-1.0-Instruct-2.3B
Microsoft/Text / LLMRuns anywhere

Microsoft's Aion 1.0 Instruct is a compact 2.3B chat model

A small, multilingual instruction-tuned model aimed at chat and summarization lands on Hugging Face.

Nov 18, 2026
SenseNova-Loop-9B
SenseTime/Vision-LanguageRuns on a laptop

SenseTime Releases Looped 9B Multimodal Model

The company's new SenseNova-Loop-9B pairs vision-language capabilities with a recurrent looped-transformer design.

Nov 17, 2026
Hunyuan Hy4-flash
Tencent/Vision-LanguageMulti-GPU server

Tencent releases Hunyuan Hy4-flash multimodal MoE

The lightweight, mixture-of-experts model adds vision and text instruction-following to Tencent's open Hunyuan lineup.

Nov 17, 2026