llama.cpp before b8585 contains a use-after-free vulnerability in the RPC server's GRAPH_RECOMPUTE handler that allows unauthenticated remote attackers to achieve arbitrary read and write access by storing a computation graph, freeing referenced buffers, and reclaiming freed memory with attacker-controlled content. Attackers can send RPC requests to trigger re-execution of stored graphs with dangling pointers, enabling full remote code execution without requiring authentication or user interaction.
CVE-2026-39909
This high-severity CVE scores 8.1 under the CNA's CVSS (NVD's own analysis pending). EPSS exploit-prediction score not yet available (the EPSS model rescores nightly; freshly-published CVEs typically appear within 48 hours). GitHub Security Advisory data not yet ingested — confidence will rise once GHSA publishes (typical lag: hours to days for open-source ecosystem CVEs; never for infrastructure-only CVEs).
- High severity, but no confirmed exploitation yet
No vendor fix yet — apply a workaround or compensating control (WAF / firewall / segmentation) and watch for a patch.
- CVSS v3
- 8.1
- EG Score
- 8.1(low)
- EG Risk
- 41(Track)EG Risk 41/100SSVC: Track
EG Risk is EchelonGraph's 0–100 priority score: it fuses intrinsic severity with real-world exploitation and automatability so you can rank equal-severity CVEs and fix the most dangerous first. Higher = act sooner. Distinct from the 0–10 EG Score (severity).
How it’s computedSeverity81% × 45%Exploitation0% × 40%Automatability30% × 15%Action: Routine — remediate on your standard cadence. - EPSS PROB
- —
- EPSS %ILE
- —
- KEV
- Not listed
Published
August 21, 2026
Last Modified
August 21, 2026
Advisory Details (4)
Auto-updated Aug 21, 2026llama.cpp Use-After-Free in RPC GRAPH_RECOMPUTE Handler | Advisories | VulnCheck
https://www.vulncheck.com/advisories/llama-cpp-use-after-free-in-rpc-graph-recompute-handlercommit 389c7d4955ba (ggml-org/llama.cpp)
Fix landed in ggml-org/llama.cpp commit 389c7d4955ba — awaiting tagged release
https://github.com/ggml-org/llama.cpp/commit/389c7d4955ba55c7418afaebf7c23d9ed64ef707webui: Fix branching logic on edit message
Fix merged in ggml-org/llama.cpp PR #21175 on 2026-03-30 — awaiting tagged release
https://github.com/ggml-org/llama.cpp/pull/21175b8585
Patch available: ggml-org/llama.cpp b8585
https://github.com/ggml-org/llama.cpp/releases/tag/b8585Weakness Classification(1)
MITRE Common Weakness Enumeration — the root-cause categories this CVE belongs to.
Data Freshness Timeline
(refreshed 2× in last 7d / 2× in last 30d)
Each row is a source pipeline that fetched or updated this CVE on that date, with what changed. For example, "NVD update" means NVD published or revised its analysis for this CVE; "MITRE cvelistV5" means we ingested or refreshed it from the CNA feed. Most recent first.
- 2026-08-21 16:51 UTCEG score recompute
- 2026-08-21 16:49 UTCMITRE cvelistV5first tracked
Related CVEs(same CWE)
Frequently asked(4)
What is CVE-2026-39909?
When was CVE-2026-39909 disclosed?
What is the CVSS score of CVE-2026-39909?
How do I remediate CVE-2026-39909?
Dependency Blast Radius
Explore the affected products and dependency analysis for CVE-2026-39909
Is Your Infrastructure Affected by CVE-2026-39909?
EchelonGraph automatically scans your cloud infrastructure and maps CVE exposure using blast radius analysis.