Huawei releases MACaQ, its internal quantizer for local inference
thefox96
81 points
23 comments
November 11, 2026
Related Discussions
Found 5 related stories in 127.9ms across 12,324 title embeddings via pgvector HNSW
- H3-metal – Native MiniMax-H3 inference for Apple Silicon swyx · 172 pts · August 11, 2026 · 53% similar
- Apple Silicon and macOS VMs: Faster LLM Inference with llama.cpp frabonacci · 289 pts · August 11, 2026 · 53% similar
- M5 Ultra Mac Studio Review piotrgrabowski · 238 pts · September 21, 2026 · 51% similar
- Qwen3.8-27B on the iPhone 18 Neural Engine at 15–70 tok/s (1.5 bpw) anemll · 30 pts · October 17, 2026 · 50% similar
- LFM2.5-DSpark: Up to 3.2x Faster Inference from H100 to MacB Alephinitesimal · 15 pts · August 21, 2026 · 49% similar
Discussion Highlights (4 comments)
utternerd
What a project name.
Rabbit504030201
https://huawei-csl.com/blog/macaq/ Why does some HN users seem to prefer GitHub page whenever available at expense of official site? I mean, Huawei is still sanctioned, anything they host third party outside China can disappear at any hour.
mocchibird
Did they release any model? Seems good for Strata
lukasc-ch
People are reproducing it and creating GGUFs https://www.reddit.com/r/LocalLLaMA/s/36atJ2nNjh