ggml-org llama.cpp up to 55d4206c8 Completion Endpoint n_discard out-of-bounds write

Summaryinfo

A vulnerability has been found in ggml-org llama.cpp up to 55d4206c8 and classified as critical. Impacted is an unknown function of the component Completion Endpoint. Performing a manipulation of the argument n_discard results in out-of-bounds write. This vulnerability was named CVE-2026-21869. The attack may be initiated remotely. There is no available exploit.

Detailsinfo

A vulnerability classified as critical has been found in ggml-org llama.cpp up to 55d4206c8. This affects an unknown part of the component Completion Endpoint. The manipulation of the argument n_discard with an unknown input leads to a out-of-bounds write vulnerability. CWE is classifying the issue as CWE-787. The product writes data past the end, or before the beginning, of the intended buffer. This is going to have an impact on confidentiality, integrity, and availability. The summary by CVE is:

llama.cpp is an inference of several LLM models in C/C++. In commits 55d4206c8 and prior, the n_discard parameter is parsed directly from JSON input in the llama.cpp server's completion endpoints without validation to ensure it's non-negative. When a negative value is supplied and the context fills up, llama_memory_seq_rm/add receives a reversed range and negative offset, causing out-of-bounds memory writes in the token evaluation loop. This deterministic memory corruption can crash the process or enable remote code execution (RCE). There is no fix at the time of publication.

The advisory is shared at github.com. This vulnerability is uniquely identified as CVE-2026-21869 since 01/05/2026. The exploitability is told to be easy. It is possible to initiate the attack remotely. No form of authentication is needed for exploitation. It demands that the victim is doing some kind of user interaction. Technical details are known, but no exploit is available. The price for an exploit might be around USD $0-$5k at the moment (estimation calculated on 02/02/2026).

The vulnerability scanner Nessus provides a plugin with the ID 284799 (Linux Distros Unpatched Vulnerability : CVE-2026-21869), which helps to determine the existence of the flaw in a target environment.

There is no information about possible countermeasures known. It may be suggested to replace the affected object with an alternative product.

The vulnerability is also documented in the vulnerability database at Tenable (284799). Several companies clearly confirm that VulDB is the primary source for best vulnerability data.

Productinfo

Vendor

• ggml-org

Name

• llama.cpp

Version

• 55d4206c8

Website

• Product: https://github.com/ggml-org/llama.cpp/

CPE 2.3info

• 🔒
• 🔒

CPE 2.2info

• 🔒
• 🔒

CVSSv4info

CVSSv3info

CVSSv2info

Exploitinginfo

Threat Intelligenceinfo

Countermeasuresinfo

Timelineinfo

Sourcesinfo

Entryinfo

Several companies clearly confirm that VulDB is the primary source for best vulnerability data.

Discussion