CVE-2026-34159: Ggml Llama.cpp

Critical severity, CVSS 9.8. EPSS: 1.2% chance of exploitation in the next 30 days.

llama.cpp is an inference of several LLM models in C/C++. Prior to version b8492, the RPC backend's deserialize_tensor() skips all bounds validation when a tensor's buffer field is 0. An unauthenticated attacker can read and write arbitrary process memory via crafted GRAPH_COMPUTE messages. Combined with pointer leaks from ALLOC_BUFFER/BUFFER_GET_BASE, this gives full ASLR bypass and remote code execution. No authentication required, just TCP access to the RPC server port. This issue has been patched in version b8492.

Affected products

• Ggml Llama.cpp: before b8492 (fixed in b8492)

Published 2026-04-01. Last modified 2026-06-17.