CVE-2026-73555
Received Received - Intake

Information Disclosure in vLLM Prior to 0.26.0

Publication date: 2026-08-13

Last updated on: 2026-08-13

Assigner: GitHub, Inc.

Description
vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the validation_exception_handler in vllm/entrypoints/openai/server_utils.py converts FastAPI RequestValidationError objects with str(exc), and sanitize_message in vllm/entrypoints/utils.py does not remove traceback-style file paths, allowing unauthenticated malformed JSON requests to /v1/chat/completions, /v1/completions, /tokenize, and /detokenize to disclose the OS username, home and virtual-environment paths, Python version, internal package structure, line numbers, and endpoint handler names. This issue is fixed in version 0.26.0.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published
2026-08-13
Last Modified
2026-08-13
Generated
2026-08-13
AI Q&A
2026-08-13
EPSS Evaluated
N/A
NVD
CVE-2026-73555
EUVD
EUVD-2026-58053

Affected Vendors & Products

Showing 2 associated CPEs
Vendor Product Version / Range
vllm_project vllm 0.26.0
vllm_project vllm to 0.26.0 (exc)

Helpful Resources

Exploitability

CWE
CWE Icon
KEV
KEV Icon
CWE ID Description
CWE-209 The product generates an error message that includes sensitive information about its environment, users, or associated data.

Attack-Flow Graph

AI Quick Actions

Instant insights powered by AI
Executive Summary

This vulnerability in vLLM versions before 0.26.0 allows unauthenticated attackers to disclose sensitive system information through error messages. When a malformed JSON request is sent to endpoints like /v1/chat/completions, FastAPI raises a validation error. The error handler converts this to a string, exposing internal file paths, line numbers, OS username, home directory, Python version, and package structure. The sanitize_message function fails to remove these details, enabling information leakage.

Detection Guidance

To detect this vulnerability, check if your vLLM version is below 0.26.0. Use commands like 'pip show vllm' or 'vllm --version' to verify the installed version. Monitor API error responses for sensitive information like file paths, usernames, or Python versions in error messages from endpoints like /v1/chat/completions.

Impact Analysis

Attackers could use disclosed system details to craft targeted exploits, identify software versions, or gain insights into server configuration. This may lead to further attacks like path traversal or privilege escalation. The leaked information could also be used for reconnaissance before launching more sophisticated attacks.

Compliance Impact

This vulnerability could violate GDPR's data protection principles by exposing personal data (e.g., OS username, home directory) and HIPAA's security requirements by leaking sensitive system information. Organizations may face compliance violations, fines, or reputational damage due to unauthorized data disclosure.

Mitigation Strategies

Upgrade vLLM to version 0.26.0 or later immediately. If upgrading is not possible, modify the validation_exception_handler in server_utils.py to avoid using str(exc) and enhance sanitize_message to strip file paths and tracebacks from error responses.

Chat Assistant

Ask questions about this CVE
Hi! I’m here to help you understand CVE-2026-73555. Ask me anything about the vulnerability, its impact, or mitigation strategies.
0/70

EPSS Chart