LLM EXPLORER 61,060 MODELS INDEXED

Whittle Next 27B A3B by logic65

By logic65 · 3427 downloads
HF Repo ↗ Copy URL

Whittle Next 27B A3B is an open-source language model by logic65. Features: 35b LLM, VRAM: 4.5GB, Context: 256K, License: apache-2.0.

Base model:finetune:qwen/qwen3... Base model:qwen/qwen3.6-35b-a3...   Code   Conversational   En   Endpoints compatible   Hyper-connections   Knowledge-distillation   Llama.cpp   Long-context   Math   Mixture-of-experts   Model-index   Moe   Multilingual   N-gram-memory   Online-distillation   Per-layer-embeddings   Qwen3   Qwen4 exp   Qwen4 exp text   Reasoning   Region:us   Research   Safetensors   Thinking   Whittle   Zh
Model Card on HF 🤗: https://huggingface.co/logic65/Whittle-Next-27B-A3B 

Whittle Next 27B A3B Parameters and Internals

LLM NameWhittle Next 27B A3B
Repository 🤗https://huggingface.co/logic65/Whittle-Next-27B-A3B 
Base Model(s)  Qwen/Qwen3.6-35B-A3B   logic65/Whittle-Next-26B-A3B   Qwen/Qwen3.6-35B-A3B   logic65/Whittle-Next-26B-A3B
Model Size35b
Required VRAM4.5 GB
Updated2026-09-14
Maintainerlogic65
Model Typeqwen4_exp_text
Model Files  4.7 GB   4.6 GB   4.6 GB   4.6 GB   4.6 GB   4.6 GB   4.6 GB   4.6 GB   4.6 GB   4.6 GB   4.5 GB   4.9 GB   0.0 GB
Supported Languagesen zh
Model ArchitectureQwen4ExpForCausalLM
Licenseapache-2.0
Context Length262144
Model Max Length262144
Transformers Version4.57.1
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size248320
Errorsreplace