The open‑source llama.cpp library released version b11496 on Oct 8 2026, introducing a SYCL‑based bulk model loading path via a pinned ring buffer.
NoteBy OossaPublished by Oossa: 1 min read
The ggml‑org team published llama.cpp b11496 on Oct 8 2026. The update adds a SYCL stage that moves bulk model uploads through a pinned ring buffer, speeding up loading on compatible GPUs. Binaries for macOS (Apple Silicon and Intel), iOS, and various Ubuntu builds (CPU, Vulkan, CUDA 12) are available for download.
Developers can load large language models faster on SYCL‑compatible hardware, reducing startup time for apps that use llama.cpp.
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.
