LLM EXPLORER 58,608 MODELS INDEXED

OpenCUA 7B by xlangai

By xlangai · 9988 downloads
HF Repo ↗ Copy URL

OpenCUA 7B is an open-source language model by xlangai. Features: 7b LLM, VRAM: 17.5GB, Context: 125K, License: mit, Instruction-Based, LLM Explorer Score: 0.24.

  Arxiv:2504.07981   Arxiv:2508.09123 Base model:finetune:qwen/qwen2... Base model:qwen/qwen2.5-vl-7b-...   Computer-use-agent   Conversational   Custom code   Dataset:osunlp/uground-v1-data Dataset:smolagents/aguvis-stag...   Dataset:xlangai/agentnet   Dataset:xlangai/aguvis-stage1   En   Feature-extraction   Grounding   Gui   Image-text-to-text   Instruct   Opencua   Os-agent   Region:us   Safetensors   Sharded   Tensorflow   Vlm
Model Card on HF 🤗: https://huggingface.co/xlangai/OpenCUA-7B 

OpenCUA 7B Parameters and Internals

LLM NameOpenCUA 7B
Repository 🤗https://huggingface.co/xlangai/OpenCUA-7B 
Base Model(s)  Qwen/Qwen2.5-VL-7B-Instruct   Qwen/Qwen2.5-VL-7B-Instruct
Model Size7b
Required VRAM17.5 GB
Updated2026-08-08
Maintainerxlangai
Model Typeopencua
Instruction-BasedYes
Model Files  2.9 GB: 1-of-28   0.5 GB: 10-of-28   0.5 GB: 11-of-28   0.5 GB: 12-of-28   0.5 GB: 13-of-28   0.5 GB: 14-of-28   0.5 GB: 15-of-28   0.5 GB: 16-of-28   0.5 GB: 17-of-28   0.5 GB: 18-of-28   0.5 GB: 19-of-28   0.5 GB: 2-of-28   0.5 GB: 20-of-28   0.5 GB: 21-of-28   0.5 GB: 22-of-28   0.5 GB: 23-of-28   0.5 GB: 24-of-28   0.5 GB: 25-of-28   0.5 GB: 26-of-28   0.5 GB: 27-of-28   1.6 GB: 28-of-28   0.5 GB: 3-of-28   0.5 GB: 4-of-28   0.5 GB: 5-of-28   0.5 GB: 6-of-28   0.5 GB: 7-of-28   0.5 GB: 8-of-28   0.5 GB: 9-of-28
Supported Languagesen
Model ArchitectureOpenCUAForConditionalGeneration
Licensemit
Context Length128000
Model Max Length128000
Transformers Version4.48.3
Tokenizer ClassTikTokenV3
Padding Token[PAD]
Vocabulary Size152064
Torch Data Typebfloat16