← Vulnerability feed

Vulnerability record · CVE-2026-61539 · published 21 August 2026

CVE-2026-61539: Xinference is an inference API for running open-source, speech, and multimodal models.

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.

10.0 CVSS 3.1 Critical EPSS 1.2% · top 32.4% CWE-95 · CWE-95 Received
10.0CVSS 3.1 base score
1.2%EPSS exploitation probability, 30 days
NoNot in CISA KEV
0Affected product versions listed by NVD
4References
24 Aug 2026Last modified by NVD

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.

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H

References

LinkTags
https://github.com/xorbitsai/inference/commit/1b3d220f342ce68d34cec4586d9409d457dadc42
https://github.com/xorbitsai/inference/pull/4786
https://github.com/xorbitsai/inference/releases/tag/v2.7.0
https://github.com/xorbitsai/inference/security/advisories/GHSA-x2rj-828p-hx9m

Track CVE-2026-61539 inside VULONE

Watch it alongside the ransomware crews, C2 infrastructure and forum chatter that reference it, query it through the API and pull it into your SIEM over TAXII.

Start free Open in platform

Related vulnerabilities

Same products first, then exploited flaws of the same weakness class.

9.3CVE-2026-33017Langflow build_public_tmp endpoint unauthenticated remote code executionLangflow versions prior to 1.9.0 expose the POST /api/v1/build_public_tmp/{flow_id}/flow endpoint without authentication, and when the optional data …KEVEPSS 25%analysed9.8CVE-2025-24893XWiki SolrSearch unauthenticated remote code executionXWiki Platform's SolrSearch endpoint evaluates user-supplied search text as Groovy code, allowing arbitrary remote code execution. The flaw is reacha…KEVEPSS 100%analysed9.8CVE-2024-36401GeoServer OGC request parameter XPath eval injection enables unauthenticated RCEGeoServer versions before 2.22.6, 2.23.6, 2.24.4, and 2.25.2 unsafely evaluate OGC request parameters as XPath expressions via the GeoTools commons-j…KEVEPSS 100%analysed7.8CVE-2023-7101Spreadsheet::ParseExcel Perl module code injection via Excel number format stringsSpreadsheet::ParseExcel 0.65, a Perl module for parsing Excel files, passes unvalidated input from a file into a string-type eval. Specifically, Numb…KEVEPSS 19%analysed

Source: NIST National Vulnerability Database (record CVE-2026-61539), CISA KEV, FIRST EPSS (scores of 2026-09-26). This page is refreshed as NVD updates the record.