LLM EXPLORER 55,709 MODELS INDEXED

Holo1 7B Bf16 by mlx-community

By mlx-community · 24 downloads
HF Repo ↗ Copy URL

Holo1 7B Bf16 is an open-source language model by mlx-community. Features: 7b LLM, VRAM: 16.6GB, Context: 125K, License: apache-2.0, Instruction-Based.

  Action   Agent Base model:finetune:qwen/qwen2... Base model:qwen/qwen2.5-vl-7b-...   Conversational   En   Endpoints compatible   Image-text-to-text   Instruct   Mlx   Multimodal   Qwen2 5 vl   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/mlx-community/Holo1-7B-bf16 

Holo1 7B Bf16 Parameters and Internals

LLM NameHolo1 7B Bf16
Repository 🤗https://huggingface.co/mlx-community/Holo1-7B-bf16 
Base Model(s)  Qwen/Qwen2.5-VL-7B-Instruct   Qwen/Qwen2.5-VL-7B-Instruct
Model Size7b
Required VRAM16.6 GB
Updated2026-08-08
Maintainermlx-community
Model Typeqwen2_5_vl
Instruction-BasedYes
Model Files  5.3 GB: 1-of-4   5.3 GB: 2-of-4   4.9 GB: 3-of-4   1.1 GB: 4-of-4
Supported Languagesen
Model ArchitectureQwen2_5_VLForConditionalGeneration
Licenseapache-2.0
Context Length128000
Model Max Length128000
Transformers Version4.51.3
Padding Token<|endoftext|>
Vocabulary Size152064
Errorsreplace

Best Alternatives to Holo1 7B Bf16

Best Alternatives
Context / RAM
Downloads
Likes
OmniLong Qwen2.5 VL 7B512K / 16.6 GB72
Qwen2.5 VL 7B Instruct FP4125K / 7.2 GB491653
Qwen2.5 VL 7B Instruct NVFP4125K / 7.2 GB809216
Qwen2.5 VL 7B Instruct FP8125K / 10 GB10098
HuatuoGPT Vision 7B Qwen2.5VL125K / 16.6 GB35811
OlmOCR 2 7B 1025 Bf16125K / 16.6 GB1542
Qwen2.5 VL 7B Instruct Bf16125K / 16.6 GB1603
ZwZ 7B125K / 16.6 GB3210
ViLaSR125K / 16.7 GB3818
GUI Owl 7B32K / 16.7 GB69853
Note: green Score (e.g. "73.2") means that the model is better than mlx-community/Holo1-7B-bf16.