@jun_song: In few weeks, everyone with 128gb Mac will have uncensored Opus-4.6 locally. It will be Minimax-M3.0-JANGTQ-CRACK by @d…
Summary
The tweet claims that an uncensored version of Opus 4.6, derived from Minimax-M3.0 and created by alignnai, will soon run locally on 128GB Macs and 24GB VRAM GPUs.
View Cached Full Text
Cached at: 05/13/26, 10:18 AM
In few weeks, everyone with 128gb Mac will have uncensored Opus-4.6 locally.
It will be Minimax-M3.0-JANGTQ-CRACK by @dealignai
The open-source community is working hard on fitting them into 24GB VRAM.
The future of Local LLM is so bright.
Similar Articles
@dealignai: MiniMax m3, made for 128gb Mac’s Thank you to @hornsby_andrew for preparing the pruning calibration dataset and doing e…
A pruned and quantized version of MiniMax-M3 (MiniMax-M3-Medium-JANG_2L) optimized to run on 128GB Macs using vMLX, featuring 32% expert pruning and JANG_2L mixed-precision quantization to fit within ~105 GB.
Given how good Qwen become, is it time to grab a 128gb m5 max?
User considers upgrading to 128GB M5 Max to run improved Qwen 27B models locally, noting near-Opus-4.5-level performance.
@dealignai: Qwen3.6-27b and 35b MXFP4 MXFP8 CRACK is out now with MTP. Enjoy uncensored speediness! 35b mxfp4: https://huggingface.…
DealignAI releases CRACK-abliterated and MXFP4/MXFP8 quantized versions of Qwen3.6-27B and 35B models, preserving MTP for faster speculative decoding on Apple Silicon.
@jundotkim: oMLX 0.3.9rc1 released. Highlights: - Low-memory Macs stay stable instead of getting killed by the OS - DFlash bumped t…
oMLX 0.3.9rc1, an LLM inference server optimized for Apple Silicon Macs, adds low-memory stability, chunked prefill, multi-tasking admin chat, and more.
JANGQ-AI/MiniMax-M2.7-JANGTQ_K : mixed-bit quant of MiniMax M2.7 - 74 GB on disk
Release of a mixed-bit quantized version of the MiniMax M2.7 model, optimized to 74 GB for efficient local inference on Apple Silicon devices.