LLM EXPLORER 59,598 MODELS INDEXED

Qwen3.8 27B PrismaScout AQUA 20GB by rdtand

By rdtand · 266 downloads
HF Repo ↗ Copy URL

Qwen3.8 27B PrismaScout AQUA 20GB is an open-source language model by rdtand. Features: 27b LLM, VRAM: 20.2GB, Context: 256K, License: apache-2.0, LLM Explorer Score: 0.33.

  8-bit   Aqua Base model:quantized:qwen/qwen...   Base model:qwen/qwen3.8-27b   Blackwell   Compressed-tensors   Conversational   Dgx-spark   Fp8   Mixed-precision   Mtp   Nvfp4   Prismaaqua   Prismaquant   Prismascout   Quantized   Qwen3.8   Qwen3 5 text   Region:us   Safetensors   Sharded   Speculative-decoding   Tensorflow   Text-only   Vllm
Model Card on HF 🤗: https://huggingface.co/rdtand/Qwen3.8-27B-PrismaScout-AQUA-20GB 

Qwen3.8 27B PrismaScout AQUA 20GB Parameters and Internals

LLM NameQwen3.8 27B PrismaScout AQUA 20GB
Repository 🤗https://huggingface.co/rdtand/Qwen3.8-27B-PrismaScout-AQUA-20GB 
Base Model(s)  Qwen/Qwen3.8-27B   Qwen/Qwen3.8-27B
Model Size27b
Required VRAM20.2 GB
Updated2026-08-26
Maintainerrdtand
Model Typeqwen3_5_text
Model Files  1.3 GB: 1-of-19   0.0 GB: 2-of-19   2.5 GB: 3-of-19   1.1 GB: 4-of-19   1.1 GB: 5-of-19   1.1 GB: 6-of-19   1.1 GB: 7-of-19   1.1 GB: 8-of-19   1.1 GB: 9-of-19   1.0 GB: 10-of-19   1.0 GB: 11-of-19   1.0 GB: 12-of-19   1.1 GB: 13-of-19   1.1 GB: 14-of-19   1.1 GB: 15-of-19   1.1 GB: 16-of-19   1.0 GB: 17-of-19   1.0 GB: 18-of-19   0.4 GB: 19-of-19
Model ArchitectureQwen3_5ForCausalLM
Licenseapache-2.0
Context Length262144
Model Max Length262144
Transformers Version5.8.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size248320
Errorsreplace

Best Alternatives to Qwen3.8 27B PrismaScout AQUA 20GB

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.6 27B OBLITERATED256K / 53.6 GB15137184
Barcenas Qwen3.8 27B Fable256K / 54.1 GB4344
Qwen3.8 27B Abliteratex256K / 53.9 GB3216
...maAQUA Gridbook 13GB 5080 Vllm256K / 13 GB5064
...wen3.8 P44w75 16.8B Unrepaired256K / 2.4 GB9291
Focus Red256K / 53.8 GB4161
...2 Abliterated OptiQ 3.7bpw Mlx256K / 13.4 GB672714
...2 Abliterated OptiQ 3.7bpw Mlx256K / 13.4 GB591914
Qwen3.6 27B Esper4256K / 55.6 GB29219
Ornstein 3.6 27B RYS256K / 56.8 GB3971
Note: green Score (e.g. "73.2") means that the model is better than rdtand/Qwen3.8-27B-PrismaScout-AQUA-20GB.