b11110 b11111
ggml-org/llama.cppSep 22, 2026 → Sep 22, 20260 days between releases
AI comparison unavailable for this pair right now — see the full release notes below for what each version contains.
The Two Releases
b11110
Pre-release
b11110
Sep 22, 2026· 30 days ago
Adds various sanity checks to the Multi-Threaded Multi-Device (MTD) implementation to improve operational robustness.
Full release notesb11111
Pre-release
b11111
Sep 22, 2026· 30 days ago
Introduces optimized Flash Attention kernels for Intel Xe GPUs (LPG Plus, Xe2, Xe3) and fixes compatibility issues with A770 Linux drivers.
Full release notes