Mac GLM-5.3-Flash Dual-ANE pack: stock Vontra oQ4 + native MTP on oMLX 0.6.3rc2. Not abliterated.
Please donate / support: https://t.co/5O4WUxexXa
- Mac GLM-5.3 Dual-ANE — oMLX 0.6.3rc2 Dual-ANE+CPU prefill for
glm5_next, Lightning MTP (layer-45 NEXTN), stock Vontra oQ4-MTP (~173 GiB). Decode 26.4 tok/s; uncached ~4k TTFT ~10% faster vs GPU.
https://github.com/drowzeys/keys-Mac-oMLX-0.6.3.2RC-Dual-ANE-GLM-5.3-Flash-oQ4
Hub card: https://huggingface.co/drowzeys/keys-Mac-oMLX-0.6.3.2RC-Dual-ANE-GLM-5.3-Flash-oQ4
Same keys-Mac-oMLX-0.6.3.2RC family as the Qwen and DSV4 Dual-ANE Mac packs.
Spark cousin (NVFP4, 2× DGX Spark, 1M ctx):
https://github.com/drowzeys/keys-vLLm.0.27.1-GLM-5.3-Flash-NVFP4-NVFP4KV-1M-Context-Abliterated
Earlier 1.0 Beta drop: 08/23/2026