CVE-2026-110695

Deferred Deferred - Pending Action

Buffer Overflow in ExLlamaV2 Safetensors Loader

Publication date: 2026-10-14

Last updated on: 2026-10-15

Assigner: VulDB

Description

A vulnerability was detected in turboderp ExLlamaV2 up to 0.3.2. Affected is the function STFile.get_tensor of the file exllamav2/stloader.py of the component Safetensors Tensor Loader. Performing a manipulation of the argument data_offsets results in buffer overflow. The attack is possible to be carried out remotely. The exploit is now public and may be used. The project was informed of the problem early through an issue report but has not responded yet.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published

2026-10-14

Last Modified

2026-10-15

Generated

2026-10-19

AI Q&A

2026-10-15

EPSS Evaluated

2026-10-18

NVD

CVE-2026-110695

EUVD

EUVD-2026-101663

Affected Vendors & Products

Vendor	Product	Version / Range
turboderp	ExLlamaV2	0.3.0
turboderp	ExLlamaV2	0.3.1
turboderp	ExLlamaV2	0.3.2

Helpful Resources

Exploitability

CWE

KEV

CWE ID	Description
CWE-120	The product copies an input buffer to an output buffer without verifying that the size of the input buffer is less than the size of the output buffer.
CWE-119	The product performs operations on a memory buffer, but it reads from or writes to a memory location outside the buffer's intended boundary. This may result in read or write operations on unexpected memory locations that could be linked to other variables, data structures, or internal program data.

Attack-Flow Graph

Executive Summary

This vulnerability is a buffer overflow in the STFile.get_tensor function of the Safetensors Tensor Loader component in turboderp ExLlamaV2 up to version 0.3.2. It occurs when the data_offsets argument is manipulated, leading to memory corruption. The issue is remotely exploitable and has a public exploit available. The vendor has not responded to the reported issue.

Detection Guidance

This vulnerability involves a buffer overflow in the STFile.get_tensor function of turboderp ExLlamaV2 up to version 0.3.2. Detection requires checking for vulnerable versions of ExLlamaV2 and monitoring for unusual memory access patterns or crashes in applications using the Safetensors Tensor Loader component.

Impact Analysis

If exploited, this vulnerability could allow an attacker to execute arbitrary code or cause a denial of service on systems running vulnerable versions of turboderp ExLlamaV2. The impact depends on the privileges of the application and the environment where it is deployed.

Compliance Impact

The vulnerability involves a buffer overflow in the STFile.get_tensor function of turboderp ExLlamaV2, which could allow remote attackers to execute arbitrary code or cause denial of service. This may lead to unauthorized access to sensitive data, potentially violating GDPR (data protection) and HIPAA (healthcare data privacy) compliance if exploited.

Mitigation Strategies

Immediately update ExLlamaV2 to a version beyond 0.3.2 if available. If no update exists, disable or restrict access to the Safetensors Tensor Loader component in exllamav2/stloader.py. Monitor for unusual network activity or buffer overflow attempts targeting the data_offsets argument.

Hi! I’m here to help you understand CVE-2026-110695. Ask me anything about the vulnerability, its impact, or mitigation strategies.

0/70