LLM EXPLORER 58,975 MODELS INDEXED

Llm Jp 4 32B A3b Thinking 8bit by mlx-community

By mlx-community · 71 downloads
HF Repo ↗ Copy URL

Llm Jp 4 32B A3b Thinking 8bit is an open-source language model by mlx-community. Features: 32b LLM, VRAM: 34.3GB, Context: 64K, License: apache-2.0, Quantized, LLM Explorer Score: 0.24.

  8-bit   8bit Base model:llm-jp/llm-jp-4-32b... Base model:quantized:llm-jp/ll...   Conversational   En   Ja   Mlx   Quantized   Qwen3 moe   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/mlx-community/llm-jp-4-32b-a3b-thinking-8bit 

Llm Jp 4 32B A3b Thinking 8bit Parameters and Internals

LLM NameLlm Jp 4 32B A3b Thinking 8bit
Repository 🤗https://huggingface.co/mlx-community/llm-jp-4-32b-a3b-thinking-8bit 
Base Model(s)  Llm Jp 4 32B A3b Thinking   llm-jp/llm-jp-4-32b-a3b-thinking
Model Size32b
Required VRAM34.3 GB
Updated2026-08-10
Maintainermlx-community
Model Typeqwen3_moe
Model Files  5.4 GB: 1-of-7   5.2 GB: 2-of-7   5.2 GB: 3-of-7   5.2 GB: 4-of-7   5.2 GB: 5-of-7   5.2 GB: 6-of-7   2.9 GB: 7-of-7
Supported Languagesen ja
Quantization Type8bit
Model ArchitectureQwen3MoeForCausalLM
Licenseapache-2.0
Context Length65536
Model Max Length65536
Transformers Version4.51.0
Tokenizer ClassTokenizersBackend
Padding Token<|endoftext|>
Vocabulary Size196608
Torch Data Typebfloat16

Best Alternatives to Llm Jp 4 32B A3b Thinking 8bit

Best Alternatives
Context / RAM
Downloads
Likes
Llm Jp 4 32B A3b Thinking64K / 64.3 GB322536
Llm Jp 4 32B A3b Base64K / 64.3 GB4435
Note: green Score (e.g. "73.2") means that the model is better than mlx-community/llm-jp-4-32b-a3b-thinking-8bit.