LLM EXPLORER 55,709 MODELS INDEXED

UI TARS 72B SFT Bf16 by mlx-community

By mlx-community · 32 downloads
HF Repo ↗ Copy URL

UI TARS 72B SFT Bf16 is an open-source language model by mlx-community. Features: 72b LLM, VRAM: 147.7GB, Context: 32K, License: apache-2.0.

  Conversational   En   Endpoints compatible   Gui   Image-text-to-text   Mlx   Multimodal   Qwen2 vl   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/mlx-community/UI-TARS-72B-SFT-bf16 

UI TARS 72B SFT Bf16 Parameters and Internals

LLM NameUI TARS 72B SFT Bf16
Repository 🤗https://huggingface.co/mlx-community/UI-TARS-72B-SFT-bf16 
Model Size72b
Required VRAM147.7 GB
Updated2026-08-08
Maintainermlx-community
Model Typeqwen2_vl
Model Files  5.2 GB: 1-of-28   5.3 GB: 2-of-28   5.3 GB: 3-of-28   5.3 GB: 4-of-28   5.3 GB: 5-of-28   5.3 GB: 6-of-28   5.3 GB: 7-of-28   5.3 GB: 8-of-28   5.3 GB: 9-of-28   5.3 GB: 10-of-28   5.3 GB: 11-of-28   5.3 GB: 12-of-28   5.3 GB: 13-of-28   5.3 GB: 14-of-28   5.3 GB: 15-of-28   5.3 GB: 16-of-28   5.3 GB: 17-of-28   5.3 GB: 18-of-28   5.3 GB: 19-of-28   5.3 GB: 20-of-28   5.3 GB: 21-of-28   5.3 GB: 22-of-28   5.3 GB: 23-of-28   5.3 GB: 24-of-28   5.3 GB: 25-of-28   5.3 GB: 26-of-28   5.3 GB: 27-of-28   4.7 GB: 28-of-28
Supported Languagesen
Model ArchitectureQwen2VLForConditionalGeneration
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.49.0
Padding Token<|endoftext|>
Vocabulary Size152064
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to UI TARS 72B SFT Bf16

Best Alternatives
Context / RAM
Downloads
Likes
QVQ 72B Preview Bf16125K / 147.7 GB233
UI TARS 72B DPO Bf1632K / 147.7 GB301
QVQ 72B Preview 3bit125K / 32.2 GB215
QVQ 72B Preview 4bit125K / 41.3 GB207
QVQ 72B Preview 6bit125K / 59.7 GB342
QVQ 72B Preview 8bit125K / 77.8 GB223
Qwen2 VL 72B Instruct 4bit32K / 41.3 GB725
Qwen2 VL 72B Instruct 8bit32K / 77.8 GB794