CVE-2026-85668: high-severity vulnerability in xorbitsai inference
Xinference 3.3.0 Unauthenticated Arbitrary-Path File Read via /v1/models/llm/auto-register
Published
21Vexday Risk Score
No sign of exploitation. No public exploitation artifact known so far.
ssvc Trackcvss 8.7epss 0.6%
exploitation probability
0.6%top 51% of all CVEs
observed exploitation
nono source reports it
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:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N
Affected products
xorbitsai · inferenceRelated CVEs — xorbitsai inference
In the same product, most dangerous first.
CVE-2026-61539CRITICALXinference: Remote code execution via unsafe `eval()` in Llama3 tool-call parsingEPSS 1.2%CVE-2026-76841HIGHXinference through 2.11.0 Remote Code Execution via Hardcoded trust_remote_code in Model LoadersEPSS 1.0%CVE-2026-112051HIGHXinference through 4.0.1 Path Traversal via model_name in Custom Model RegistrationEPSS 0.4%
References
https://github.com/xorbitsai/inferencehttps://github.com/xorbitsai/inference/blob/v3.3.0/xinference/model/llm/config_parser.pyhttps://github.com/xorbitsai/inference/issues/5176https://www.vulncheck.com/advisories/xinference-3.3.0-unauthenticated-arbitrary-path-file-read-via-v1-models-llm-auto-register