[Release] SOTA GGUFs for Qwen3.8-Flash-Next: GSQ-RCO Providing Near Baseline Performance

The Qwen3.8-Flash-Next model, specifically the GSQ-RCO-GGUF variant, offers a Q2_0 quantization format designed for speed. This format achieves 3.4x the prompt throughput and 1.9x lower end-to-end latency compared to IQ2_XS, with a slightly smaller file size. While it has a task average of 89.07, slightly lower than IQ2_XS (89.16) and IQ3_XXS, its decode rate remains consistent across workloads, making it ideal when throughput is the primary concern.