A vulnerability was found in ggml-org llama.cpp bec4772f6. (CVE-2026-78147)
A vulnerability exists in the ggml-org llama.cpp project affecting the deserialize_tensor function in ggml/src/ggml-rpc/ggml-rpc.cpp within the ggml-RPC Server component. The flaw involves improper handling of the op/op_params arguments during deserialization, which can be triggered remotely. This issue is distinct from a previously reported vulnerability (CVE-2026-34159) and was closed due to inactivity. No patch or fix information is currently available.
AI Analysis
Technical Summary
The vulnerability in ggml-org llama.cpp (CVE-2026-78147) affects the deserialize_tensor function of the ggml-RPC Server component. It arises from manipulation of the op/op_params arguments leading to unsafe deserialization. The vulnerability can be exploited remotely and is categorized under CWE-20 (Improper Input Validation). It is separate from CVE-2026-34159, which addressed a different issue. The GitHub issue related to this vulnerability was closed automatically due to inactivity, and no patch or remediation details have been provided.
Potential Impact
Successful exploitation could allow an attacker to cause limited confidentiality, integrity, and availability impacts as indicated by the CVSS vector (C:L/I:L/A:L). The vulnerability is remotely exploitable without privileges or user interaction. However, no known exploits are reported in the wild.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Since no patch or official fix information is available, monitor vendor channels for updates. Avoid exposing the vulnerable service to untrusted networks if possible until a fix is released.
A vulnerability was found in ggml-org llama.cpp bec4772f6. (CVE-2026-78147)
Description
A vulnerability exists in the ggml-org llama.cpp project affecting the deserialize_tensor function in ggml/src/ggml-rpc/ggml-rpc.cpp within the ggml-RPC Server component. The flaw involves improper handling of the op/op_params arguments during deserialization, which can be triggered remotely. This issue is distinct from a previously reported vulnerability (CVE-2026-34159) and was closed due to inactivity. No patch or fix information is currently available.
CVSS v3.1
Score 7.3high
Weaknesses
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
The vulnerability in ggml-org llama.cpp (CVE-2026-78147) affects the deserialize_tensor function of the ggml-RPC Server component. It arises from manipulation of the op/op_params arguments leading to unsafe deserialization. The vulnerability can be exploited remotely and is categorized under CWE-20 (Improper Input Validation). It is separate from CVE-2026-34159, which addressed a different issue. The GitHub issue related to this vulnerability was closed automatically due to inactivity, and no patch or remediation details have been provided.
Potential Impact
Successful exploitation could allow an attacker to cause limited confidentiality, integrity, and availability impacts as indicated by the CVSS vector (C:L/I:L/A:L). The vulnerability is remotely exploitable without privileges or user interaction. However, no known exploits are reported in the wild.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Since no patch or official fix information is available, monitor vendor channels for updates. Avoid exposing the vulnerable service to untrusted networks if possible until a fix is released.
Technical Details
- Gcve Source
- db.gcve.eu
- Osv Id
- GHSA-f4vj-w5xq-5xph
- Osv Schema Version
- 1.4.0
- Aliases
- ["CVE-2026-78147"]
- Ecosystems
- []
- Database Specific Severity
- MODERATE
- Cvss Version
- 3.1
Threat ID: 6a8c4c7dacd9273b499c104a
Added to database: 08/24/2026, 13:51:57 UTC
Last enriched: 08/24/2026, 14:32:12 UTC
Last updated: 08/24/2026, 16:51:58 UTC
Views: 6
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.