Unsloth releases Qwen3.8-27B quantized: NVFP4 1.5x faster, retains 92-97% accuracy
danielhanchen · x · 2026-08-14
Unsloth announced NVFP4 and dynamic GGUF quantizations for Qwen3.8-27B. NVFP4 is 1.5x faster than BF16 while retaining 92-97% top-1% accuracy. UD-IQ2XXS retains 82.5% accuracy at 9GB, 83.5% smaller than BF16 (54.7GB). These quantized models run locally on 17GB RAM, making Qwen3.8-27B the strongest model for its size that can run locally.
Related event: Alibaba Open-Sources Native Multimodal Qwen3.8-27B Model(48 posts)→
More from Infra
- TokenSpeed MLA 2 hits NVIDIA Vera Rubin: up to 3.44x faster decode vs B200 — zhyncs42 · 2026-10-16
- US vs Chinese hyperscalers: Americans fund and schmooze, Chinese demand your BoM and cap you at 7% margin — jwt0625 · 2026-10-16
- How LiteLLM routes one OpenAI-shaped API across every model provider — Mahmoud_Zalt · 2026-10-16
- OpenAI-Broadcom custom AI chip Jalapeño named a TIME Best Invention of 2026 — itsclivetime · 2026-10-16
- Developer live-runs continual pretraining of a Luganda LLM on Colab, 41% done with loss dropping fast — weskambale · 2026-10-16
- Jensen Huang: Nvidia charges ~$thousands per GPU yearly for software, a major revenue line — rohanpaul_ai · 2026-10-16