CVE-2026-61539Public exploitNetwork attack vector

Remote code execution via unsafe Llama3 tool-call evaluation

Published Aug 21, 2026 · Updated Aug 21, 2026

Eval injection in XorbitsAI Xinference 2.5.0 and earlier allows remote attackers to execute code through crafted chat-completion prompts. The Llama3 tool-call parsers in llama3_tool_parser.py and utils.py pass request-influenced, model-generated Python expressions to eval() during non-streaming result post-processing. A request must reach /v1/chat/completions with a tools field and use the Transformers path; the tested default deployment requires no authentication, and injected code runs as the server process.

CVSS severity10.0
Critical
EPSS probability0.66%
Next 30 days · Sep 16, 2026
Known exploitationUnconfirmed
Based on sourced intelligence
Hinoki checkNot available
Coverage for this vulnerability

See if you're affected

Explore vulnerability checks for your environment with Hinoki.

Book a demo

Summary

What happened

Eval injection in XorbitsAI Xinference 2.5.0 and earlier allows remote attackers to execute code through crafted chat-completion prompts. The Llama3 tool-call parsers in llama3_tool_parser.py and utils.py pass request-influenced, model-generated Python expressions to eval() during non-streaming result post-processing. A request must reach /v1/chat/completions with a tools field and use the Transformers path; the tested default deployment requires no authentication, and injected code runs as the server process.

The record

CVE
CVE-2026-61539
Published
Aug 21, 2026
Updated
Aug 21, 2026
Vendor
XorbitsAI
Product
Xinference
Classifications
CWE-95, T1190
Attack vector
network
Privileges
unauthenticated

Timeline

How it unfolded

  1. Aug 21, 2026CVE publishedPublication date reported by the CVE source.
  2. Aug 21, 2026Record updatedLatest update available in the CVE record.

Exploitability

Present is not the same as exploitable

Compare your product and version with the public record. A matching version still requires validation against your environment.

Is a vulnerable build present?

Compare these published version ranges with your installed build and any vendor patches.

  1. Affected versionversion=< 2.7.0

What conditions does exploitation require?

Attack vectornetwork
Required privilegesunauthenticated

What is affected?

XorbitsAI · Xinferenceversion=< 2.7.0

Published CVSS scores

10.0GitHubCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H

CVSS describes severity. EPSS estimates exploitation probability.

Attacks

What attackers are doing with it

Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.

Daily unique IPsNo honeypot observations are available for this CVE in the selected window.

No observations available

Sep 9, 2026Sep 15, 2026
Latest reporting daySep 15, 2026
Latest daily unique IPsUnavailable
Prior 30-day averageUnavailable
SourceShadowserver honeypots (KEV)
Vectornetwork
Privilegesunauthenticated
Known exploitationUnconfirmed
Public exploitPublished

Weakness, pattern, technique

CWE-95Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection')
T1190Exploit Public-Facing Application

Public exploit references

Labels summarize the accepted research assessment. They do not indicate a test against your environment.

Technologies

Your stack

See the directory against your own environment.

Your stack

Check the software in your environment

Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.

Book a demo