Qwen 3.8 Swift 1.5 Flash Heretic GGUF Quantizations Released

Published 31 October 2026

An uncensored 177B-parameter GGUF model optimized for coding, reverse engineering, agentic tool use, and spatial reasoning. IQ3_S, IQ4_XS-plus, and IQ4_NL variants range from 84.6 GB to 109 GB for large-VRAM deployments.