PatchSiren

PatchSiren cyber security CVE debrief

CVE-2026-86288 ModelCloud CVE debrief

A vulnerability was found in ModelCloud GPTQModel up to 7.2.0, affecting the Triton dequantization kernel in the file gptqmodel/nn_modules/qlinear/tritonv2.py. This issue allows for an out-of-bounds read due to manipulation of the argument g_idx, which can be exploited remotely. The exploit has been disclosed publicly. Upgrading to version 7.3.0 resolves this issue, with the patch being 877c732f7d7dccd56a729844c6a5bd20f3aa8bb1.

Vendor
ModelCloud
Product
GPTQModel
CVSS
LOW 2.1
CISA KEV
Not listed in stored evidence
Original CVE published
2026-09-07
Original CVE updated
2026-09-07
Advisory published
2026-09-07
Advisory updated
2026-09-07

Who should care

Defenders responsible for ModelCloud GPTQModel deployments should assess the vulnerability and prioritize upgrading to version 7.3.0 if affected. This includes operators, platform administrators, vulnerability management teams, and security teams who need to verify exposure, apply the patch, and monitor for potential exploitation attempts.

Why it matters

CVE-2026-86288 is a vulnerability in ModelCloud GPTQModel that allows for remote exploitation leading to out-of-bounds read. Defenders should verify exposure, prioritize patching, and monitor for exploitation attempts.

  • Verify exposure of ModelCloud GPTQModel versions up to 7.2.0 in the environment.
  • Assess the need for and apply the patch 877c732f7d7dccd56a729844c6a5bd20f3aa8bb1.
  • Monitor for potential exploitation attempts targeting this vulnerability.

Technical summary

The vulnerability affects ModelCloud GPTQModel up to version 7.2.0, specifically in the Triton dequantization kernel within the file gptqmodel/nn_modules/qlinear/tritonv2.py. An out-of-bounds read can occur due to manipulation of the g_idx argument. This issue can be exploited remotely. The exploit has been disclosed publicly. Upgrading to version 7.3.0 resolves this issue, with the patch being 877c732f7d7dccd56a729844c6a5bd20f3aa8bb1. Defenders should prioritize verifying exposure of ModelCloud GPTQModel versions up to 7.2.0 and assess the need for upgrading to version 7.3.0.

Defensive priority

Defenders should prioritize verifying exposure of ModelCloud GPTQModel versions up to 7.2.0 and assess the need for upgrading to version 7.3.0.

Recommended defensive actions

  • Verify if ModelCloud GPTQModel versions up to 7.2.0 are in use and assess the need for upgrading to version 7.3.0.
  • Apply the patch 877c732f7d7dccd56a729844c6a5bd20f3aa8bb1 to affected systems.
  • Monitor for potential exploitation attempts targeting this vulnerability.
  • Review compensating controls for exposed systems while remediation is scheduled and verified.
  • Check relevant monitoring, detection, and logs for exposed assets that need extra review.
  • Track exceptions, retest remediated assets, and close the item only after evidence is documented.
  • Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up.

Evidence notes

The CVE record and source references provide details on the vulnerability, its impact, and the available patch. However, the corpus does not establish versions, exploitation, impact, or remediation beyond upgrading to version 7.3.0.

Sources and references

Verified primary and authoritative sources

  • CVE-2026-86288 CVE Program record

    Publisher, destination, and source semantics verified

    URL: https://www.cve.org/CVERecord?id=CVE-2026-86288

    CVE Program - Official CVE Program record with source-provided CVE metadata.

  • CVE-2026-86288 NVD vulnerability detail

    Publisher, destination, and source semantics verified

    URL: https://nvd.nist.gov/vuln/detail/CVE-2026-86288

    NIST National Vulnerability Database - Official NIST NVD detail page and source-specific vulnerability assessment.

Supplemental references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.