Qwen3.8-Flash-Next-GGUF

All projects

Apps & Tools

Qwen3.8-Flash-Next-GGUF

Experimental open-weight hybrid attention model with GGUF quants by Unsloth

byUnsloth AI

Playable

Qwen3.8-Flash-Next-GGUF media is blocked

Allow external media to connect to the provider and play this content.

Description

Qwen3.8-Flash-Next is an experimental preview of the architecture for Qwen4, featuring a rethinking of core LLM components. It introduces Hybrid Attention with QSA (Qwen Sparse Attention) and Gated DeltaNet to reduce long-context latency, alongside n-gram embeddings for efficient parameter scaling.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.

Details

Why this is interesting: Showcases a novel hybrid attention and n-gram embedding architecture designed to underpin the upcoming Qwen4 generation.
Attributed to: @unsloth
Built with: Qwen3.8-Flash-Next Qwen3.8-27B Gemini 3.7 Flash Qwen3.7 Flash 3.6 Flash 3.5 Flash Cyber Step 3.7 Flash Gemini 3.5 Flash 2.5 Flash Qwen3 14B Qwen3 30B-A3B Qwen3 8B
Links: huggingface.co
View tweet
Tags: #llm #quantization #qwen #unsloth

Discovered: X
First posted: X · Aug 26, 2026
Listed: Added 2h ago
Build: Strong
Models: Strong