CVE-2026-61539

CRITICALCVSS 10/10EPSS 0.66%

Last modified

CVE-2026-61539 is a critical-severity vulnerability rated 10/10 on the CVSS scale. 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. EPSS estimates a 0.66% chance of exploitation in the next 30 days.

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.

Metrics

EPSS Probability
0.66%

48.7th percentile

Probability of exploitation in the next 30 days. Learn more

Weakness Enumeration

Affected Software

Source: CNA advisory (CVE.org). NVD analysis pending.

VendorProductVersions
xorbitsaiinference< 2.7.0

References

Timeline

Published
Last Modified
Status
Received

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 score of 10/10 (CRITICAL severity). The EPSS model estimates a 0.66% probability of exploitation in the next 30 days.
How do I fix CVE-2026-61539?
Check the vendor references and advisories linked above for patched versions and mitigation guidance. You can also run a Strix scan to test if your systems are affected.

How Strix Helps

Related CVEs from 2026

Are you affected by CVE-2026-61539?

Run a free Strix scan to check your systems for this vulnerability.

Scan your code now

Source: NVD / NIST