Read the docs
TI-2026-69141 Aliases CVE-2026-107183GHSA-3g9g-qh2j-c5q8

Use-after-free in llama.cpp chat parser enables heap corruption

Priority MEDIUM 42CWE-416Published · Updated

Key facts

Attack vector
Network
Privileges required
None
User interaction
None
Attack complexity
High
Attack requirements
None
Impact type
Code execution
Confidentiality · integrity · availability
High · High · High
Fix status
Fixed

Attack profile from CVSS 4.0 · CNA. Fix status from CNA. Impact type from CWE-416 and the CVSS impact.

Summary

llama.cpp from b8227 before b11393 contains a use-after-free and double free in common_chat_peg_mapper::map, caused by a dangling current_tool pointer (CWE-416). An unauthenticated remote attacker can corrupt heap memory by sending a crafted chat_parser in a POST /completion request. The same request can crash llama-server and shape a heap write primitive.

Written with Claude Haiku 5.5 from the sources below · 2026-10-29

Description from the CNA

llama.cpp before b11393 contains a use-after-free and double free vulnerability in common_chat_peg_mapper::map that allows unauthenticated remote attackers to corrupt heap memory via a dangling current_tool pointer. Attackers can submit a chat_parser in a POST /completion request emitting a tool-id after a tool-close tag to crash llama-server and shape a heap write primitive.

Why it matters now

Exploitation None reported CISA SSVC · not in CISA KEV
EPSS 0.00417 36th percentileScored 2026-11-07
Fix Fixed llama.cpp (ggml-org): b11393

SSVC decision points

CISA ADP · SSVC 2.0.3

Exploitation
None
Automatable
No
Technical impact
Total

The decision (Track, Attend, Act) also depends on your mission and well-being impact; see CISA's SSVC guide.

Affected products

VendorProductAffected versionsFixed inStatusSource
ggml-orgllama.cppb8227 to before b11393b11393Patch availableCNA
  • llama.cpp Vendorggml-org Affected b8227 to before b11393 Fixed in b11393 Patch available CNA

CPE names Threatist (inferred) · 1

Inferred from how NVD named this product before; not given by a source.

CPEVersionsRoleVendor
cpe:2.3:a:ggml:llama.cpp:*:*:*:*:*:*:*:*——ggml-org
  • cpe:2.3:a:ggml:llama.cpp:*:*:*:*:*:*:*:* Versions — — ·ggml-org

Remediation

Update llama.cpp to fixed version b11393.

Written with Claude Haiku 5.5 from the sources below · 2026-10-29

  • Fix · CNA llama.cpp (ggml-org) Fixed in b11393

Vendor assessments

Red Hat

Red Hat's page for this CVE →

Not affected: 1 product
  • Red Hat

Timeline

  1. 2026-10-07 Threatist record created from CVE-2026-107183
  2. 2026-10-07 CISA SSVC decision points changed
  3. 2026-10-07 Priority changed: MEDIUM → LOW
  4. 2026-10-07 Alias added: GHSA-3g9g-qh2j-c5q8
  5. 2026-10-08 Affected versions changed · Red Hat
  6. 2026-10-08 CVSS score added · Red Hat
  7. 2026-10-09 Priority changed: LOW → MEDIUM

References

Patch 2

Other 3

  • TI-2026-26459 llama.cpp: Unauthenticated RCE via GRAPH_COMPUTE buffer=0 bypass in llama.cpp RPC backend CVSS 3.1CRITICAL 9.8
  • TI-2024-28446 llama.cpp allows Arbitrary Address Read in rpc_server::get_tensor CVSS 3.1CRITICAL 9.8
  • TI-2026-17570 llama.cpp has Out-of-bounds Write in llama-server CVSS 3.1CRITICAL 9.8
  • TI-2024-28447 llama.cpp allows write-what-where in rpc_server::set_tensor CVSS 3.1CRITICAL 10.0
  • TI-2024-15627 Integer overflow in ggml-org llama.cpp CVSS 3.1CRITICAL 9.8
  • TI-2024-14482 Heap-based buffer overflow in ggml-org llama.cpp CVSS 3.1CRITICAL 9.8
See all 52 →

Sources

This product uses the NVD API but is not endorsed or certified by the NVD. Sources and licences