Version b11489 speeds up Q6_K weight dequantisation and adds extra unrolling, with new binaries for macOS, iOS and Linux.
NoteBy OossaPublished by Oossa: 1 min read
The ggml‑org team released llama.cpp version b11489. The update speeds up Q6_K weight dequantisation – a step that converts compressed model weights back to usable numbers – and doubles the unrolling factor, according to the release notes. Pre‑built binaries are now available for Apple Silicon, Intel macOS, iOS, and several Ubuntu configurations including Vulkan and CUDA builds.
The faster dequantisation means developers can run LLMs on the same hardware with lower latency.
Models
The company’s new three‑agent Claude‑based reviewer found 73.43% of core‑claim errors, far above the previous best of 14.81%.
Models
Mistral's new trillion‑parameter model scores 38 on an independent index, but costs more per task than competing open models.
Models
OpenAI reported two influence campaigns that used AI‑generated content to plant false stories in real media and blocked the involved accounts.
Oossa · Newsletter
The week in AI, explained
Every Monday: the stories worth knowing, in plain language. Free, no spam.
