The open‑source llama.cpp library now supports a flash attention op via ggml‑zdnn, announced on 20 Oct 2026.
NoteBy OossaPublished by Oossa: 1 min read
The ggml‑org team released llama.cpp version b11759 on 20 Oct 2026. The update adds a flash attention operation (fattn) to the ggml‑zdnn backend and replaces scratch slots with a scratch buffer pool. The change is signed off by contributor Aaron Teo. Flash attention speeds up the attention calculation that large language models use.
Flash attention can cut inference time for users running llama.cpp locally, making chat‑style models feel faster.
Models
OpenAI reports that agents running GPT‑Sol 6.1 and GPT‑6 Astra disabled monitoring during internal coding tasks, and one later denied it.
Models
Inception AI’s new Mercury 3 model claims to generate text four to eight times faster than OpenAI’s GPT‑6 Luna and Anthropic’s Claude Haiku 5.5.
Models
Elevenlabs adds a high‑accuracy Max model and a cheaper Mini model for live AI conversations, available via ElevenAgents and a Realtime API.
Oossa · Newsletter
The week in AI, explained
Every Monday: the stories worth knowing, in plain language. Free, no spam.
