LLM EXPLORER 60,958 MODELS INDEXED

Qwopus3.8 27B Flash Mlx OQ4e 1M by Solstice-AI

By Solstice-AI · 211 downloads
HF Repo ↗ Copy URL

Qwopus3.8 27B Flash Mlx OQ4e 1M is an open-source language model by Solstice-AI. Features: 27b LLM, VRAM: 16.9GB, Context: 1024K, License: apache-2.0.

  1m-context   27b   4-bit Base model:jackrong/qwopus3.8-... Base model:quantized:jackrong/...   Conversational   En   Image-text-to-text   Long-context   Mlx   Mmproj   Multimodal   Qwen   Qwen3.8   Qwen3 5   Qwopus   Region:us   Safetensors   Sharded   Solstice-ai   Tensorflow   Vision   Yarn   Zh
Model Card on HF 🤗: https://huggingface.co/Solstice-AI/Qwopus3.8-27B-Flash-mlx-oQ4e-1M 

Qwopus3.8 27B Flash Mlx OQ4e 1M Parameters and Internals

LLM NameQwopus3.8 27B Flash Mlx OQ4e 1M
Repository 🤗https://huggingface.co/Solstice-AI/Qwopus3.8-27B-Flash-mlx-oQ4e-1M 
Model Name/mnt/w/Qwen3.8-27B-Brainwaves-NM-HERETIC
Base Model(s)  Jackrong/Qwopus3.8-27B-Flash   Jackrong/Qwopus3.8-27B-Flash
Model Size27b
Required VRAM16.9 GB
Updated2026-09-13
MaintainerSolstice-AI
Model Typeqwen3_5
Model Files  5.0 GB: 1-of-4   5.0 GB: 2-of-4   5.0 GB: 3-of-4   1.9 GB: 4-of-4
Supported Languagesen zh
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Context Length1048576
Model Max Length1048576
Tokenizer ClassTokenizersBackend
Padding Token<|endoftext|>
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to Qwopus3.8 27B Flash Mlx OQ4e 1M

Best Alternatives
Context / RAM
Downloads
Likes
Agens Pilot1024K / 54.7 GB1321
Qwen3.6 27B0K / 55.4 GB67281382197
Qwen3.8 27B MLX0K / 16 GB141118
Qwen3.8 27B0K / 55.6 GB272721
Qwen3.8 27B FP80K / 0.4 GB123157416
Qwen3.8 27B0K / 55.6 GB28753
Qwen3.6 27B NVFP40K / 22 GB1887525419
Qwen3.8 27B NVFP40K / 22 GB1919574
Qwen3.6 27B FP80K / 0.4 GB7740472342
Cortex Mini 1 Preview0K / 55.6 GB1252
Note: green Score (e.g. "73.2") means that the model is better than Solstice-AI/Qwopus3.8-27B-Flash-mlx-oQ4e-1M.