CVE-2026-75090
Received Received - Intake

Out-of-Bounds Read in Mistral.rs GGUF Tokenizer

Publication date: 2026-08-18

Last updated on: 2026-08-18

Assigner: VulDB

Description
A vulnerability was detected in EricLBuehler Mistral.rs up to 0.8.22. Affected by this issue is the function convert_gguf_to_hf_tokenizer of the file mistralrs-core/src/gguf/gguf_tokenizer.rs of the component GGUF Tokenizer. The manipulation of the argument eos_token_id/bos_token_id/unknown_token_id results in out-of-bounds read. The attack can be executed remotely. The exploit is now public and may be used. Upgrading to version 0.8.23 can resolve this issue. The patch is identified as cd5297e2ea5cb27c790bdcf2f3c2f1064a81d55e. Upgrading the affected component is recommended.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published
2026-08-18
Last Modified
2026-08-18
Generated
2026-08-18
AI Q&A
2026-08-18
EPSS Evaluated
N/A
NVD
CVE-2026-75090
EUVD
EUVD-2026-60564

Affected Vendors & Products

Showing 2 associated CPEs
Vendor Product Version / Range
ericlbuehler mistral.rs to 0.8.22 (inc)
ericlbuehler mistral.rs 0.8.23

Helpful Resources

Exploitability

CWE
CWE Icon
KEV
KEV Icon
CWE ID Description
CWE-125 The product reads data past the end, or before the beginning, of the intended 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

AI Quick Actions

Instant insights powered by AI
Executive Summary

This vulnerability is an out-of-bounds read issue in the EricLBuehler Mistral.rs library up to version 0.8.22. It occurs in the GGUF Tokenizer component where special token IDs like eos_token_id, bos_token_id, and unknown_token_id are not validated against the vocabulary size before use. This allows maliciously crafted GGUF files to cause an out-of-bounds read, potentially leading to a denial-of-service condition when loading models.

Detection Guidance

To detect this vulnerability, check if you are running mistral.rs versions up to 0.8.22. Inspect GGUF model files for out-of-bounds token IDs (eos_token_id, bos_token_id, unknown_token_id) by examining metadata. Look for crashes during model loading, especially with malformed GGUF files.

Impact Analysis

The vulnerability can be exploited remotely by attackers who craft malicious GGUF files with out-of-bounds token IDs. When loaded, these files cause the application to crash with a panic (denial-of-service) during model initialization. This affects systems using affected versions of mistral.rs for LLM inference, potentially disrupting services that rely on these models.

Compliance Impact

This vulnerability does not directly affect compliance with GDPR or HIPAA. It is a denial-of-service issue in the mistral.rs library that could cause crashes when loading malformed GGUF files. No evidence suggests it impacts data protection or privacy requirements under these regulations.

Mitigation Strategies

Upgrade mistral.rs to version 0.8.23 or later. Apply the patch cd5297e2ea5cb27c790bdcf2f3c2f1064a81d55e. Validate all GGUF files for proper token ID bounds before loading models to prevent out-of-bounds access.

Chat Assistant

Ask questions about this CVE
Hi! I’m here to help you understand CVE-2026-75090. Ask me anything about the vulnerability, its impact, or mitigation strategies.
0/70

EPSS Chart