Xinference: Remote code execution via unsafe `eval()` in Llama3 tool-call parsing
Summary
| CVE | CVE-2026-61539 |
|---|---|
| State | PUBLISHED |
| Assigner | GitHub_M |
| Source Priority | CVE Program / NVD first with legacy fallback |
| Published | 2026-08-21 21:17:00 UTC |
| Updated | 2026-08-21 21:17:00 UTC |
| Description | Xinference is an inference API for running open-source, speech, and multimodal models. In 2.5.0 and earlier, Xinference passes attacker-influenced Llama3 tool-call output to eval() in xinference/model/llm/tool_parsers/llama3_tool_parser.py and xinference/model/llm/utils.py. Requests to /v1/chat/completions with a tools field flow through xinference/api/restful_api.py, xinference/model/llm/transformers/core.py, handle_chat_result_non_streaming(), and _post_process_completion() before extract_tool_calls() or _eval_llama3_chat_arguments() evaluates the model-generated Python expression. An unauthenticated remote attacker can influence that output through a crafted prompt and execute commands in the Xinference server process context. This issue is fixed in version 2.7.0. |
Risk And Classification
Primary CVSS: v3.1 10 CRITICAL from [email protected]
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H
Problem Types: CWE-95 | CWE-95 CWE-95: Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection')
| Version | Source | Type | Score | Severity | Vector |
|---|---|---|---|---|---|
| 3.1 | [email protected] | Secondary | 10 | CRITICAL | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H |
| 3.1 | CNA | DECLARED | 10 | CRITICAL | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H |
CVSS v3.1 Breakdown
Attack Vector
NetworkAttack Complexity
LowPrivileges Required
NoneUser Interaction
NoneScope
ChangedConfidentiality
HighIntegrity
HighAvailability
HighCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H
Vendor Declared Affected Products
References
| Reference | Source | Link | Tags |
|---|---|---|---|
| github.com/xorbitsai/inference/security/advisories/GHSA-x2rj-828p-hx9m | [email protected] | github.com | |
| github.com/xorbitsai/inference/releases/tag/v2.7.0 | [email protected] | github.com | |
| github.com/xorbitsai/inference/commit/1b3d220f342ce68d34cec4586d9409d457... | [email protected] | github.com | |
| github.com/xorbitsai/inference/pull/4786 | [email protected] | github.com | |
| CVE Program record | CVE.ORG | www.cve.org | canonical |
| NVD vulnerability detail | NVD | nvd.nist.gov | canonical, analysis |
No vendor comments have been submitted for this CVE.
There are currently no legacy QID mappings associated with this CVE.