The open‑source LLM runtime released new binaries with 128/96 flash‑attention support for Apple Silicon on Oct 9 2026.
NoteBy OossaPublished by Oossa: 1 min read
The ggml‑org team updated llama.cpp (release b11518) on Oct 9 2026. The change adds 128‑bit and 96‑bit flash‑attention kernels for Metal, speeding up inference on Apple Silicon Macs. Pre‑built binaries for macOS (arm64), macOS Intel, iOS, and several Linux configurations are provided for download.
Apple‑silicon users can run LLMs faster without buying extra GPU hardware.
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.
