CVE-2026-34159 llama.cpp: Unauthenticated RCE via GRAPH_COMPUTE buffer=0 bypass in llama.cpp RPC backend

What is CVE-2026-34159?

CVE-2026-34159 is a critical memory corruption vulnerability in ggml-org llama.cpp, published on 1 Apr 2026. An attacker can exploit it over the internet without logging in, with no action from the victim.

Official description

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.

Weakness: CWE-119

Which versions are affected by CVE-2026-34159?

It affects 1 product.

Products include Ggml Llama.cpp.

How do I fix CVE-2026-34159?

Upgrade to a fixed version.

Upgrade to the fixed version. Memory-safe languages, compiler hardening and fuzzing prevent most new cases. More about Memory corruption.

Is CVE-2026-34159 being exploited?

It is not confirmed as exploited in the wild, but public exploit code for CVE-2026-34159 exists.

Public exploit code: 3 proofs of concept on GitHub.

Other vulnerabilities in ggml-org llama.cpp

• CVE-2024-42479 llama.cpp allows write-what-where in rpc_server::set_tensor Critical 10.0
• CVE-2026-43629 llama.cpp b4882–b9058 Buffer Overflow in KV Cache State Restore Critical 9.2
• CVE-2026-43631 llama.cpp b7492–b9060 Use-After-Free RCE via llama-server Critical 9.2
• CVE-2026-43632 llama.cpp b7492–b9060 Use-After-Free in Tokenization Endpoints Critical 9.2
• CVE-2026-21869 llama.cpp has Out-of-bounds Write in llama-server High 8.8

Sources

• CVE record
• NVD
• OSV
• GitHub advisory
• How this page is made

Also known as DEBIAN-CVE-2026-34159, GHSA-j8rj-fmpv-wcxw, UBUNTU-CVE-2026-34159.