LLM EXPLORER 58,531 MODELS INDEXED

Qwen3 32B MLX Bf16 by Qwen

By Qwen · 505 downloads
HF Repo ↗ Copy URL

Qwen3 32B MLX Bf16 is an open-source language model by Qwen. Features: 32b LLM, VRAM: 65.2GB, Context: 40K, License: apache-2.0, LLM Explorer Score: 0.2.

  Arxiv:2309.00071   Arxiv:2505.09388   Conversational   Mlx   Qwen3   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/Qwen/Qwen3-32B-MLX-bf16 

Qwen3 32B MLX Bf16 Parameters and Internals

LLM NameQwen3 32B MLX Bf16
Repository 🤗https://huggingface.co/Qwen/Qwen3-32B-MLX-bf16 
Model Size32b
Required VRAM65.2 GB
Updated2026-08-01
MaintainerQwen
Model Typeqwen3
Model Files  5.2 GB: 1-of-13   5.3 GB: 2-of-13   5.1 GB: 3-of-13   5.1 GB: 4-of-13   5.3 GB: 5-of-13   5.1 GB: 6-of-13   5.1 GB: 7-of-13   5.3 GB: 8-of-13   5.1 GB: 9-of-13   5.1 GB: 10-of-13   5.3 GB: 11-of-13   5.1 GB: 12-of-13   3.1 GB: 13-of-13
Model ArchitectureQwen3ForCausalLM
Licenseapache-2.0
Context Length40960
Model Max Length40960
Transformers Version4.51.3
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size151936
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to Qwen3 32B MLX Bf16

Best Alternatives
Context / RAM
Downloads
Likes
...ll Qwen3 32B Preview7 QAT 200K195K / 65.8 GB103
...ll Qwen3 32B Preview6 QAT 200K195K / 65.8 GB62
...ll Qwen3 32B Preview4 QAT 200K195K / 65.8 GB132
Qwen3 32B AWorld128K / 65.8 GB3715
Qwen3 32B40K / 65.8 GB541115
Qwen3 32B40K / 65.6 GB6155624715
OpenThinkerAgent 32B40K / 65.8 GB90229
Awakening Goedel V2 32B Sft40K / 65.8 GB5690
Awakening Goedel V2 32B Rl40K / 65.8 GB5650
Qwen3 Swallow 32B SFT V0.240K / 65.6 GB244110
Note: green Score (e.g. "73.2") means that the model is better than Qwen/Qwen3-32B-MLX-bf16.