GPU crash and disclosure via unchecked checkpoint index
Published Sep 7, 2026 · Updated Sep 8, 2026
Out-of-bounds read in ModelCloud GPTQModel 7.0 through 7.2.0 allows attackers to crash GPU inference via a crafted checkpoint. The Triton dequantization kernel uses checkpoint-supplied g_idx values to index scales and qzeros buffers without verifying that the resolved group is below num_groups. Exploitation requires loading and running an untrusted desc_act checkpoint; it can crash the CUDA context and expose adjacent GPU memory through dequantized weights.
Summary
What happened
Out-of-bounds read in ModelCloud GPTQModel 7.0 through 7.2.0 allows attackers to crash GPU inference via a crafted checkpoint. The Triton dequantization kernel uses checkpoint-supplied g_idx values to index scales and qzeros buffers without verifying that the resolved group is below num_groups. Exploitation requires loading and running an untrusted desc_act checkpoint; it can crash the CUDA context and expose adjacent GPU memory through dequantized weights.
The record
- CVE
- CVE-2026-86288
- Published
- Sep 7, 2026
- Updated
- Sep 8, 2026
- Vendor
- ModelCloud
- Product
- GPTQModel
- Classifications
- CWE-119, CWE-125, T1204.002
- Attack vector
- network
- Privileges
- unauthenticated
Timeline
How it unfolded
- Sep 7, 2026CVE publishedPublication date reported by the CVE source.
- Sep 8, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=7.0
- Affected versionversion=7.1
- Affected versionversion=7.2.0
What conditions does exploitation require?
What is affected?
Published CVSS scores
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- GPTQModel issue 2949 proof of conceptproof of concept · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo