LLM EXPLORER 63,287 MODELS INDEXED

Qwen3.8 Whittle 16B by logic65

By logic65 · 1719 downloads
HF Repo ↗ Copy URL

Qwen3.8 Whittle 16B is an open-source language model by logic65. Features: 27b LLM, VRAM: 2.4GB, Context: 256K, License: apache-2.0, Quantized, LLM Explorer Score: 0.29.

Base model:quantized:qwen/qwen... Base model:qwen/qwen3.8-27b-fp...   Conversational   Depth-pruning   Endpoints compatible   Gated-deltanet   Gguf   Healed   Pruning   Q4   Quantized   Qwen3 5   Region:us   Safetensors   Width-pruning
Model Card on HF 🤗: https://huggingface.co/logic65/Qwen3.8-Whittle-16B 

Qwen3.8 Whittle 16B Parameters and Internals

LLM NameQwen3.8 Whittle 16B
Repository 🤗https://huggingface.co/logic65/Qwen3.8-Whittle-16B 
Base Model(s)  Qwen/Qwen3.8-27B-FP8   Qwen/Qwen3.8-27B-FP8
Model Size27b
Required VRAM2.4 GB
Updated2026-08-18
Maintainerlogic65
Model Typeqwen3_5
Model Files  10.1 GB   17.4 GB   2.5 GB   4.0 GB   4.0 GB   3.9 GB   4.0 GB   3.9 GB   4.0 GB   4.0 GB   2.4 GB
GGUF QuantizationYes
Quantization Typeq4|gguf|q4_k
Model ArchitectureQwen3_5ForCausalLM
Licenseapache-2.0
Context Length262144
Model Max Length262144
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size248320
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to Qwen3.8 Whittle 16B

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.8 27B OBLITERATED256K / 83.2 GB164950516
JiRackDeltaNet 27b256K / 10.9 GB5559403
...Bonsai 2 27B Heretic Satellite256K / 53.9 GB41770
Qwen3.8 27B OBLITERATEDpintus256K / 83.2 GB9811
...onsai 27B Mlx Lossless 1.75bpw0K / 5.9 GB3121
...tic Uncensored NM DAU MXFP4 1M1024K / 0.1 GB15021
...tic Uncensored NM DAU NVFP4 1M256K / 0.8 GB43572
...eretic Uncensored NM DAU NVFP4256K / 0.1 GB17672
...Heretic Abliterated Uncensored256K / 53.9 GB12832
...Uncensored Heretic Abliterated256K / 53.8 GB22712
Note: green Score (e.g. "73.2") means that the model is better than logic65/Qwen3.8-Whittle-16B.