LLM EXPLORER 62,155 MODELS INDEXED

Grug 67B A2B Datakit SFT 262K 2026.09.21 by open-athena

By open-athena · 467 downloads
HF Repo ↗ Copy URL

Grug 67B A2B Datakit SFT 262K 2026.09.21 is an open-source language model by open-athena. Features: 67b LLM, VRAM: 133.9GB, Context: 256K, License: openmdw-1.1, LLM Explorer Score: 0.31.

  Conversational   Grug   Grug moe   Marin   Mixture-of-experts   Reasoning   Region:us   Safetensors   Sft   Sharded   Tensorflow   Tool-use
Model Card on HF 🤗: https://huggingface.co/open-athena/Grug-67B-A2B-Datakit-SFT-262K-2026.09.21 

Grug 67B A2B Datakit SFT 262K 2026.09.21 Parameters and Internals

LLM NameGrug 67B A2B Datakit SFT 262K 2026.09.21
Repository 🤗https://huggingface.co/open-athena/Grug-67B-A2B-Datakit-SFT-262K-2026.09.21 
Model Size67b
Required VRAM133.9 GB
Updated2026-09-25
Maintaineropen-athena
Model Typegrug_moe
Model Files  4.7 GB: 1-of-39   3.4 GB: 2-of-39   3.4 GB: 3-of-39   3.4 GB: 4-of-39   3.4 GB: 5-of-39   3.4 GB: 6-of-39   3.4 GB: 7-of-39   3.4 GB: 8-of-39   3.4 GB: 9-of-39   3.4 GB: 10-of-39   3.4 GB: 11-of-39   3.4 GB: 12-of-39   3.4 GB: 13-of-39   3.4 GB: 14-of-39   3.4 GB: 15-of-39   3.4 GB: 16-of-39   3.4 GB: 17-of-39   3.4 GB: 18-of-39   3.4 GB: 19-of-39   3.4 GB: 20-of-39   3.4 GB: 21-of-39   3.4 GB: 22-of-39   3.4 GB: 23-of-39   3.4 GB: 24-of-39   3.4 GB: 25-of-39   3.4 GB: 26-of-39   3.4 GB: 27-of-39   3.4 GB: 28-of-39   3.4 GB: 29-of-39   3.4 GB: 30-of-39   3.4 GB: 31-of-39   3.4 GB: 32-of-39   3.4 GB: 33-of-39   3.4 GB: 34-of-39   3.4 GB: 35-of-39   3.4 GB: 36-of-39   3.4 GB: 37-of-39   3.4 GB: 38-of-39   3.4 GB: 39-of-39
Model ArchitectureGrugMoeForCausalLM
Licenseopenmdw-1.1
Context Length262144
Model Max Length262144
Transformers Version5.12.1
Tokenizer ClassPreTrainedTokenizerFast
Vocabulary Size128256

Best Alternatives to Grug 67B A2B Datakit SFT 262K 2026.09.21

Best Alternatives
Context / RAM
Downloads
Likes
...2B Datakit SFT 262K 2026.09.11256K / 70.6 GB3391
...7B A2b Sft S2 Thinking Step63064K / 133.9 GB3500
...7B A2b Sft S1 Wildchat Step25764K / 133.9 GB2330