The open‑source Llama.cpp library released version b12224 on Nov 9 2026, fixing an input upload issue and providing pre‑built binaries for macOS, iOS and several Ubuntu platforms.
NoteBy OossaPublished by Oossa: 1 min read
ggml‑org released llama.cpp version b12224. The update stops sending input data that the backend never reads, fixing a minor performance bug. It also ships ready‑to‑run binaries for macOS Apple Silicon, macOS Intel, iOS, and multiple Ubuntu CPU builds, including support for s390x and Vulkan. The release notes link to the download files and attestations.
Developers can now run Llama.cpp locally on more hardware without building from source, saving setup time.
Models
Pinterest says the model improved email clicks and push opens in US and Canadian tests by reading activity histories for longer-term goals.
Models
The model is available through the company’s Tinker service. Thinking Machines says it can handle coding, research and business tasks, and plans to release its weights.
Models
The 65-billion-parameter model is available under Apache 2.0. Ricursive reports a strong coding benchmark score, but results vary across tests.
Oossa · Newsletter
The week in AI, explained
Every Monday: the stories worth knowing, in plain language. Free, no spam.
