How severe is CVE-2026-61539?

CVE-2026-61539 has a CVSS 3.x base score of 10.0, rated critical severity. It is exploitable over network with low attack complexity, requires no privileges and no user interaction. Impact on confidentiality is high, integrity high, and availability high.

Is CVE-2026-61539 being actively exploited?

It is not currently listed in CISA's KEV catalog. Its EPSS exploit-prediction score is 1% (49th percentile), an estimate of the probability of exploitation in the next 30 days.

How do I fix CVE-2026-61539?

Review the linked vendor and NVD advisories for patched versions and mitigations, then upgrade or apply the recommended workaround. Given its critical severity, prioritise patching exposed systems.

When was CVE-2026-61539 published?

CVE-2026-61539 was published on 2026-08-21.

CVE-2026-61539

CVE-2026-61539 is a critical-severity vulnerability with a CVSS 3.x base score of 10.0. It is not currently listed as actively exploited by CISA, and its EPSS exploit-prediction score is low. The underlying weakness is classified as CWE-95.

Key facts

Severity: Critical (CVSS 3.x base score 10.0)
EPSS exploit prediction: 1% (49th percentile)
Actively exploited: Not listed in CISA KEV
Weakness: CWE-95
Published: 2026-08-21
Last modified: 2026-08-21

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.

Frequently asked questions

What is CVE-2026-61539?: 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.
How severe is CVE-2026-61539?: CVE-2026-61539 has a CVSS 3.x base score of 10.0, rated critical severity. It is exploitable over network with low attack complexity, requires no privileges and no user interaction. Impact on confidentiality is high, integrity high, and availability high.
Is CVE-2026-61539 being actively exploited?: It is not currently listed in CISA's KEV catalog. Its EPSS exploit-prediction score is 1% (49th percentile), an estimate of the probability of exploitation in the next 30 days.
How do I fix CVE-2026-61539?: Review the linked vendor and NVD advisories for patched versions and mitigations, then upgrade or apply the recommended workaround. Given its critical severity, prioritise patching exposed systems.
When was CVE-2026-61539 published?: CVE-2026-61539 was published on 2026-08-21.

References

Other CWE-95 (Eval Injection) vulnerabilities

CVE-2026-44643 — Critical (CVSS 10.0): Angular Expressions provides expressions for the Angular.JS web framework as a standalone module. Prior to 1.5.2, an…
CVE-2025-68271 — Critical (CVSS 10.0): OpenC3 COSMOS provides the functionality needed to send commands to and receive data from one or more embedded systems.…
CVE-2013-10070 — Critical (CVSS 10.0): PHP-Charts v1.0 contains a PHP code execution vulnerability in wizard/url.php, where user-supplied GET parameter names…
CVE-2022-36010 — Critical (CVSS 10.0): This library allows strings to be parsed as functions and stored as a specialized component,…
CVE-2026-77810 — Critical (CVSS 9.9): In the Neptune connector, a user with access to Neptune through Athena Federated Query could gain access to properties…
CVE-2026-19626 — Critical (CVSS 9.9): A remote code execution vulnerability exists in Tenable Security Center's report generation functionality. An…

Browse all CWE-95 (Eval Injection) vulnerabilities →