Text-Only Qwen3.5 2B/4B/9B MLX 4-bit Packages Released, 2B Is Just 1GB
sachasayan · reddit · 2026-10-09
A Redditor released text-only (vision-removed) MLX 4-bit packages of Qwen3.5 for Apple Silicon local inference: 2B, 4B and 9B, each in original and Huihui abliterated variants — six packs total, with the 2B version at only 1GB. The author runs them in their writing app Minstrel for summarization and contextual decisions; downloads are on Hugging Face, crediting Qwen, Huihui and community converters.
More from Models
- Dev argues for "open-weight models" over "open-source": you can't contribute to them — kipperrii · 2026-10-09
- ChatGPT Invented Court Cases and Lawyers Got Suspended: Inside AI's Legal Hallucination Failures — dadakoglu · 2026-10-09
- User says Qwen3.8 in Hermes "hacked" his PC to prep for CPA exam — natesiggard · 2026-10-09
- Early user: GPT-6 web research feels 10x faster than 5.6 in ChatGPT — flowersslop · 2026-10-09
- FineWeb author: annotating pretraining data with a 27B model is wild but pays off at deployment — antoine_chaffin · 2026-10-09
- HF researcher: fine-tuned small models win on throughput, zero-shot wins on capabilities — antoine_chaffin · 2026-10-09