VulnerabilityPublished 7 Nov 2025Updated 15 Nov 2025

CVE-2025-64496: Open WebUI Affected by an External Model Server (Direct Connections) Code Injection via SSE Events

Affects open-webuiSource GCVE Database

Compare

Authoritative record:CVE-2025-64496· High 7.3

Not known to be exploited · CVSS 7.3 High : upgrade to 0.6.35 soon.

Data as of 13 Nov 2026, 19:05 UTC

CVSS 3.1 base score

7.3High

How severe the flaw is on the 0–10 CVSS scale.

Thin evidence

One scorer or none

CVSS scores by source
CNA (GitHub_M)	CVSS 3.1	7.3

Patch: Available – fixed in 0.6.35

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

Attack Vector: Network
Attack Complexity: Low
Privileges Required: Low
User Interaction: Required
Scope: Unchanged
Confidentiality: High
Integrity: High
Availability: None

Bold values are the worst case: easiest to exploit or largest impact.

Attack Vector: Network: The vulnerable component is bound to the network stack and the set of possible attackers extends beyond the other options listed, up to and including the entire Internet. Such a vulnerability is often termed 'remotely exploitable' and can be thought of as an attack being exploitable at the protocol level one or more network hops away (e.g., across one or more routers).
Attack Complexity: Low: Specialized access conditions or extenuating circumstances do not exist. An attacker can expect repeatable success against the vulnerable component.
Privileges Required: Low: The attacker is authorized with (i.e., requires) privileges that provide basic user capabilities that could normally affect only settings and files owned by a user. Alternatively, an attacker with Low privileges may have the ability to cause an impact only to non-sensitive resources.
User Interaction: Required: Successful exploitation of this vulnerability requires a user to take some action before the vulnerability can be exploited.
Scope: Unchanged: An exploited vulnerability can only affect resources managed by the same security authority. In this case, the vulnerable component and the impacted component are either the same, or both are managed by the same security authority.
Confidentiality: High: There is total loss of confidentiality, resulting in all resources within the impacted component being divulged to the attacker. Alternatively, access to only some restricted information is obtained, but the disclosed information presents a direct, serious impact.
Integrity: High: There is a total loss of integrity, or a complete loss of protection. For example, the attacker is able to modify any/all files protected by the impacted component. Alternatively, only some files can be modified, but malicious modification would present a direct, serious consequence to the impacted component.
Availability: None: There is no impact to availability within the impacted component.

Definitions: FIRST CVSS v3.1 specification.

Exploitation evidence

Disclosed: evidence found.
CVE published 8 Nov 2025
Detectable: no evidence on record.
No public nuclei template indexed
Weaponized: no evidence on record.
No public exploit code indexed in Metasploit or Exploit-DB
Exploited in the wildExploited: no evidence on record.
Not in CISA KEV and no source reports exploitation
Observed by OffSeq MirageSeen by Mirage: no evidence on record.
No attack traffic recorded by OffSeq Mirage sensors

A missing step is not evidence of safety. Sources: cve.org (opens in a new tab), the Metasploit, Exploit-DB and nuclei-templates index as of 13 Nov 2026.

Disclosed: The CVE record is public on cve.org. Its publication is t = 0.
Detectable: A public nuclei template can find exposed instances, for defenders and attackers alike.
Weaponized: A Metasploit exploit module or Exploit-DB entry exists (Metasploit scanner modules do not count), or CISA SSVC or the CVSS exploit maturity reports a public proof of concept.
Exploited in the wild: CISA KEV lists it, CISA SSVC rates exploitation active, or a source reports attacks.
Observed by OffSeq Mirage: OffSeq Mirage honeypot sensors recorded attack traffic for this CVE, or found exposed hosts without attack traffic.

Moments are spaced evenly, not to scale; a dashed line marks a gap of six months or more. Deltas count from t = 0. Sources that give only a date (CISA KEV, Exploit-DB, or a time of exactly 00:00Z) read “same day” rather than hours. Metasploit modules carry the flaw’s disclosure date, not their release date, and nuclei templates carry no date, so both appear above but not on the timeline.

What to do

Upgrade open-webui to 0.6.35.

Affected software

Product	Affected	Fixed in
npmopen-webui	Affected <0.6.35	Fixed in 0.6.35
PyPIopen-webui	Affected <0.6.35	Fixed in 0.6.35

Check your own hosts for these packages with threat-finder (opens in a new tab), our free open-source scanner.

AI-suggested

A patch is available in Open WebUI version 0.6.35 that addresses this vulnerability. Users should upgrade to version 0.6.35 or later to remediate the issue. Until patched, it is recommended to keep the Direct Connections feature disabled (it is disabled by default) and avoid adding untrusted external model servers. No other mitigations are specified by the vendor advisory.

Description

Open WebUI versions 0.6.33 and below have a code injection vulnerability in the Direct Connections feature. This flaw allows malicious external model servers to execute arbitrary JavaScript in users' browsers via Server-Sent Event (SSE) 'execute' events. Exploitation can lead to theft of authentication tokens, full account takeover, and, when combined with another vulnerability in the Functions API, remote code execution on the backend server. The attack requires the victim to enable Direct Connections and add the attacker's malicious server URL, typically via social engineering. A fix is available in version 0.6.35.

Impact

Successful exploitation allows attackers to execute arbitrary JavaScript in victim browsers, resulting in theft of permanent JWT authentication tokens and complete account takeover. With admin privileges obtained, attackers can leverage a chained vulnerability in the Functions API to achieve remote code execution on the backend server. This compromises both client and server security, potentially leading to full system compromise.

Summary

Open WebUI's Direct Connections feature improperly handles Server-Sent Events from external model servers without validating their trustworthiness or filtering event types. Specifically, it executes JavaScript code received in SSE 'execute' events using 'new Function()', allowing arbitrary code execution in the browser context. This enables attackers to steal JWT tokens stored in localStorage, which do not expire, leading to full account compromise. The stolen admin token can then be used to exploit the Functions API for remote code execution on the backend server. The vulnerability affects Open WebUI versions prior to 0.6.35 and was confirmed on version 0.6.33. The root cause includes lack of validation, absence of allowlisting, no sandboxing, and insufficient protection of credentials.

CWE-501Trust Boundary Violation (opens in a new tab)
Class weaknessOWASP Top 10 A06:2025: Insecure Design (opens in a new tab)
The product mixes trusted and untrusted data in the same data structure or structured message.
OWASP Cheat Sheets
Secure Design Principles (opens in a new tab)
From MITRE CWE, MITRE CAPEC and OWASP (Top 10, OpenCRE).
CWE-829Inclusion of Functionality from Untrusted Control Sphere (opens in a new tab)
Base weaknessOWASP Top 10 A08:2025: Software or Data Integrity Failures (opens in a new tab)
The product imports, requires, or includes executable functionality (such as a library) from a source that is outside of the intended control sphere.
Architecture and Design · Libraries or Frameworks
Use a vetted library or framework that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid.
Architecture and Design · Enforcement by Conversion
When the set of acceptable objects, such as filenames or URLs, is limited or known, create a mapping from a set of fixed input values (such as numeric IDs) to the actual filenames or URLs, and reject all other inputs.
Architecture and Design
For any security checks that are performed on the client side, ensure that these checks are duplicated on the server side, in order to avoid CWE-602. Attackers can bypass the client-side checks by modifying values after the checks have been performed, or by changing the client to remove the client-side checks…
OWASP Cheat Sheets
Docker Security (opens in a new tab)
Vulnerable Dependency Management (opens in a new tab)
Software Supply Chain Security (opens in a new tab)
Infrastructure as Code (opens in a new tab)
OWASP ASVS 4.0.3
V12.3.6 (opens in a new tab)
V14.2.3 (opens in a new tab)
V14.2.4 (opens in a new tab)
Attack patterns
Serialized Data External Linking (opens in a new tab)
DTD Injection (opens in a new tab)
Local Code Inclusion (opens in a new tab)
PHP Local File Inclusion (opens in a new tab)
Remote Code Inclusion (opens in a new tab)
From MITRE CWE, MITRE CAPEC and OWASP (Top 10, OpenCRE).
CWE-830Inclusion of Web Functionality from an Untrusted Source (opens in a new tab)
Variant weaknessOWASP Top 10 A08:2025: Software or Data Integrity Failures (opens in a new tab)
The product includes web functionality (such as a web widget) from another domain in a way that causes the functionality to operate within the domain of the product, potentially granting total access and control of the product to the untrusted source.
OWASP Cheat Sheets
Unvalidated Redirects and Forwards (opens in a new tab)
XML Security (opens in a new tab)
XML External Entity Prevention (opens in a new tab)
File Upload (opens in a new tab)
Input Validation (opens in a new tab)
OWASP ASVS 4.0.3
V5.3.6 (opens in a new tab)
V12.3.6 (opens in a new tab)
V14.2.3 (opens in a new tab)
From MITRE CWE, MITRE CAPEC and OWASP (Top 10, OpenCRE).
CWE-95Improper Neutralization of Directives in Dynamically Evaluated Code (Eval Injection) (opens in a new tab)
Variant weaknessLikelihood of exploit: MediumOWASP Top 10 A05:2025: Injection (opens in a new tab)
The product receives input from an upstream component, but it does not neutralize or incorrectly neutralizes code syntax before using the input in a dynamic evaluation call (e.g. "eval").
Architecture and Design · Implementation · Refactoring
If possible, refactor your code so that it does not need to use eval() at all.
Implementation · Input Validation
Assume all input is malicious. Use an "accept known good" input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does.
Implementation
Inputs should be decoded and canonicalized to the application's current internal representation before being validated (CWE-180, CWE-181). Make sure that your application does not inadvertently decode the same input twice (CWE-174).
OWASP Cheat Sheets
Deserialization (opens in a new tab)
XML Security (opens in a new tab)
XML External Entity Prevention (opens in a new tab)
Server Side Request Forgery Prevention (opens in a new tab)
Unvalidated Redirects and Forwards (opens in a new tab)
OWASP ASVS 4.0.3
V5.5.4 (opens in a new tab)
V5.2.4 (opens in a new tab)
V1.5.1 (opens in a new tab)
Attack patterns
Leverage Executable Code in Non-Executable Files (opens in a new tab)
From MITRE CWE, MITRE CAPEC and OWASP (Top 10, OpenCRE).

References

Community reviews

Loading reviews…

Exploitation evidence

What to do

Affected software

Description

AI analysisMachine-generated, updated 13 Nov 2026, 19:05 UTC

WeaknessesCWE-501, CWE-829, CWE-830 +1

Technical details

References

Community reviews

Related threats