LLM EXPLORER 67,627 MODELS INDEXED

Bonsai 2 27B W4A16 by PixelML

By PixelML · 359 downloads

Bonsai 2 27B W4A16 is an open-source language model by PixelML. Features: 27b LLM, VRAM: 16.5GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.27.

Base model:quantized:qwen/qwen...   Base model:qwen/qwen3.8-27b   Bonsai   Compressed-tensors   Conversational   Dflash2   Endpoints compatible   Gguf   Gptq   Image-text-to-text   Quantized   Qwen3 5   Region:us   Safetensors   Sharded   Speculative-decoding   Tensorflow   Ternary   W4a16

Bonsai 2 27B W4A16 Parameters and Internals

LLM NameBonsai 2 27B W4A16
Repository πŸ€—https://huggingface.co/PixelML/Bonsai-2-27B-W4A16 
Base Model(s)  Qwen/Qwen3.8-27B   prism-ml/Ternary-Bonsai-2-27B-gguf   Qwen/Qwen3.8-27B   prism-ml/Ternary-Bonsai-2-27B-gguf
Model Size27b
Required VRAM16.5 GB
Updated2026-09-22
MaintainerPixelML
Model Typeqwen3_5
Model Files  16.1 GB: 1-of-2   0.4 GB: 2-of-2   0.2 GB   0.0 GB
GGUF QuantizationYes
Quantization Typegguf
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version5.14.1
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to Bonsai 2 27B W4A16

Best Alternatives
Context / RAM
Downloads
Likes
...N V1.1 NM DAU NEO MAX MTP GGUF0K / 0.9 GB421538328
Qwen3.80K / 0.9 GB541
... Opus Reasoning Distilled GGUF0K / 0.9 GB67921679
Qwen3.8 27B GGUF0K / 0.9 GB0685
Qwen3.8 27B OBLITERATED0K / 83.2 GB25442
...B Abliterated V2 PQ2 0 MTP MLX0K / 0.8 GB19717
...N Blackwell DFlash2 Ultra V1.00K / 0.9 GB35313
Swift Qwen3.8 27B MIX GGUF0K / 0.9 GB22192
...inkingCap Qwen3.8 27B MIX GGUF0K / 13.4 GB13957
EleMo V3 Base0K / 0.9 GB19891
Note: green Score (e.g. "73.2") means that the model is better than PixelML/Bonsai-2-27B-W4A16.