LLM EXPLORER 61,135 MODELS INDEXED

Qwen3.8 Flash Next NVIDIA Hybrid Single Spark by Shinrali

By Shinrali · 463 downloads
HF Repo ↗ Copy URL

Qwen3.8 Flash Next NVIDIA Hybrid Single Spark is an open-source language model by Shinrali. Features: 118.4b LLM, VRAM: 76GB, License: other, LLM Explorer Score: 0.31.

  8-bit Base model:nvidia/qwen3.8-flas... Base model:quantized:nvidia/qw...   Conversational   Dgx-spark   Fp8   Image-text-to-text   Modelopt   Multimodal   Nvfp4   Quantization   Qwen3.8   Qwen4 exp   Region:us   Safetensors   Sharded   Tensorflow   Video   Vision   Vision-language   Vllm
Model Card on HF 🤗: https://huggingface.co/Shinrali/Qwen3.8-Flash-Next-NVIDIA-Hybrid-Single-Spark 

Qwen3.8 Flash Next NVIDIA Hybrid Single Spark Parameters and Internals

LLM NameQwen3.8 Flash Next NVIDIA Hybrid Single Spark
Repository 🤗https://huggingface.co/Shinrali/Qwen3.8-Flash-Next-NVIDIA-Hybrid-Single-Spark 
Base Model(s)  nvidia/Qwen3.8-Flash-Next-NVFP4   nvidia/Qwen3.8-Flash-Next-NVFP4
Model Size118.4b
Required VRAM76 GB
Updated2026-09-14
MaintainerShinrali
Model Typeqwen4_exp
Model Files  3.0 GB: 1-of-10   9.6 GB: 2-of-10   9.6 GB: 3-of-10   9.6 GB: 4-of-10   9.7 GB: 5-of-10   9.6 GB: 6-of-10   9.6 GB: 7-of-10   9.6 GB: 8-of-10   5.6 GB: 9-of-10   0.1 GB: 10-of-10   51.2 GB   1.6 GB
Model ArchitectureQwen4ExpForConditionalGeneration
Licenseother
Model Max Length262144
Transformers Version5.16.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace