CVE-2026-85668: External Control of File Name or Path in xorbitsai inference
Description
Xinference (affected commit 4a94832, v3.x) contains an unauthenticated arbitrary-path file read vulnerability in the POST /v1/models/llm/auto-register endpoint, which accepts a caller-supplied model_path parameter without authentication or path confinement. The endpoint reads and parses config.json, tokenizer_config.json, and chat_template.jinja files at the supplied path and reflects the parsed content back to the caller, allowing an unauthenticated attacker to probe the server filesystem and extract content of files with those names in any directory.
CVSS v4.0
Score 8.7high
Affected software
xorbitsai
inference
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
CVE-2026-85668 describes an unauthenticated arbitrary-path file read vulnerability in the xorbitsai inference product, specifically in Xinference version 3.x at commit 4a94832. The POST /v1/models/llm/auto-register endpoint accepts a model_path parameter from callers without authentication or path confinement. It reads and parses config.json, tokenizer_config.json, and chat_template.jinja files located at the supplied path and returns their parsed content to the caller. This allows an unauthenticated attacker to probe the server filesystem and extract the contents of these files from any directory.
Potential Impact
An unauthenticated attacker can read arbitrary files named config.json, tokenizer_config.json, and chat_template.jinja from any directory on the server hosting the vulnerable Xinference service. This leads to unauthorized disclosure of potentially sensitive configuration data, which could aid further attacks or information gathering. There is no indication of privilege escalation or code execution from this vulnerability alone.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until an official fix is available, restrict network access to the vulnerable endpoint and monitor for suspicious requests targeting the model_path parameter. Implement authentication and path validation controls if possible to limit exposure.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- VulnCheck
- Date Reserved
- 2026-09-04T13:32:30.391Z
- Cvss Version
- 4.0
- State
- PUBLISHED
Threat ID: 6a9adb45acd9273b49bef593
Added to database: 09/04/2026, 14:52:53 UTC
Last enriched: 09/07/2026, 12:15:49 UTC
Last updated: 10/17/2026, 18:48:23 UTC
Views: 77
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.