CVE-2026-0599VulnerabilityPublished 2 Feb 2026

CVE-2026-0599: huggingface/text-generation-inference · Uncontrolled Resource Consumption

Affects huggingface/text-generation-inferenceSource CVE Database V5

Not known to be exploited · 29% EPSS · CVSS 7.5 High : upgrade to 3.3.7 soon.

Data as of 26 Oct 2026, 06:26 UTC

CVSS 3.0 base score

7.5High

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

EPSS: 29.8% chance of exploitation in 30 days – higher than 98% of CVEsExploit Prediction Scoring System: the likelihood attackers will use this flaw soon.
Patch: Available – fixed in 3.3.7

CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

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

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: None: The attacker is unauthorized prior to attack, and therefore does not require any access to settings or files to carry out an attack.
User Interaction: None: The vulnerable system can be exploited without any interaction from any user.
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: None: There is no loss of confidentiality within the impacted component.
Integrity: None: There is no loss of integrity within the impacted component.
Availability: High: There is total loss of availability, resulting in the attacker being able to fully deny access to resources in the impacted component; this loss is either sustained (while the attacker continues to deliver the attack) or persistent (the condition persists even after the attack has completed). Alternatively, the attacker has the ability to deny some availability, but the loss of availability presents a direct, serious consequence to the impacted component (e.g., the attacker cannot disrupt existing connections, but can prevent new connections; the attacker can repeatedly exploit a vulnerability that, in each instance of a successful attack, leaks a only small amount of memory, but after repeated exploitation causes a service to become completely unavailable).

Definitions: FIRST CVSS v3.1 specification.

Exploitation evidence

Disclosed: evidence found.
CVE published 2 Feb 2026
Detectable: no evidence on record.
No public nuclei template indexed
Weaponised: 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 2026-10-26.

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.
Weaponised: Public exploit code exists in Metasploit or Exploit-DB, or CISA SSVC 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 huggingface/text-generation-inference to 3.3.7.

Affected software

Product	Affected	Fixed in
huggingfacehuggingface/text-generation-inference	Affected ≥ unspecified, < 3.3.7	Fixed in–
GitHubhuggingface/text-generation-inference`pkg:github/huggingface/text-generation-inference`Other threats	Affected >=3.3.6 <3.3.7	Fixed in 3.3.7

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

AI-suggested

European organizations should immediately upgrade huggingface/text-generation-inference to version 3.3.7 or later, where the vulnerability is fixed. Until upgrading, implement strict network egress controls to restrict outbound HTTP requests from the inference service, preventing arbitrary external image fetching. Configure resource limits such as memory caps, CPU quotas, and request timeouts at the container or orchestration level (e.g., Kubernetes resource limits) to contain potential resource exhaustion. Enable authentication and authorization mechanisms to restrict access to the inference API, especially if deployed in public or semi-public environments. Implement input validation or sanitization to detect and block Markdown image links before processing. Monitor system resource usage and network traffic for anomalous spikes indicative of exploitation attempts. Consider deploying web application firewalls (WAFs) or API gateways with rate limiting and request inspection to mitigate abuse. Finally, maintain an incident response plan to quickly isolate and remediate affected systems if exploitation is suspected.

Description

A vulnerability in huggingface/text-generation-inference version 3.3.6 allows unauthenticated remote attackers to exploit unbounded external image fetching during input validation in VLM mode. The issue arises when the router scans inputs for Markdown image links and performs a blocking HTTP GET request, reading the entire response body into memory and cloning it before decoding. This behavior can lead to resource exhaustion, including network bandwidth saturation, memory inflation, and CPU overutilization. The vulnerability is triggered even if the request is later rejected for exceeding token limits. The default deployment configuration, which lacks memory usage limits and authentication, exacerbates the impact, potentially crashing the host machine. The issue is resolved in version 3.3.7.

Impact

For European organizations deploying huggingface/text-generation-inference version 3.3.6 or earlier, this vulnerability poses a significant risk of denial-of-service (DoS) attacks. Attackers can remotely trigger resource exhaustion by submitting inputs containing Markdown image links pointing to large or slow-responding external resources, causing excessive memory and CPU consumption and network bandwidth saturation. This can lead to service outages, degraded performance, and potential crashes of critical AI inference infrastructure. Organizations relying on this software for AI-driven applications, including natural language and vision-language processing, may experience operational disruptions. The lack of authentication in default deployments increases exposure, especially for publicly accessible inference services. Such outages could impact customer-facing services, internal automation, or research environments. Additionally, the resource exhaustion could be leveraged as a smokescreen for other attacks or to degrade trust in AI services. Given the growing adoption of Hugging Face tools in Europe, the threat is material and requires prompt mitigation.

CWE-400Uncontrolled Resource Consumption (opens in a new tab)
Class weaknessLikelihood of exploit: High
The product does not properly control the allocation and maintenance of a limited resource.
MITRE discourages mapping vulnerabilities to this broad entry; the root cause is usually a more specific weakness.
Architecture and Design
Design throttling mechanisms into the system architecture. The best protection is to limit the amount of resources that an unauthorized user can cause to be expended. A strong authentication and access control model will help prevent such attacks from occurring in the first place.
Architecture and Design
Mitigation of resource exhaustion attacks requires that the target system either: recognizes the attack and denies that user further access for a given amount of time, or; uniformly throttles all requests in order to make it more difficult to consume resources more quickly than they can again be freed.
Architecture and Design
Ensure that protocols have specific limits of scale placed on them.
OWASP Cheat Sheets
File Upload (opens in a new tab)
OWASP ASVS 4.0.3
V12.1.1 (opens in a new tab)
Attack patterns
XML Ping of the Death (opens in a new tab)
Regular Expression Exponential Blowup (opens in a new tab)
Sustained Client Engagement (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 2 Feb 2026, 11:14 UTC

WeaknessesCWE-400

Technical details

References

Community reviews

Related threats