CVE-2026-114052 vllm-project CVE debrief

Who should care

Defenders responsible for vLLM installations, especially those with unauthenticated API access, should assess exposure and prioritize patching or mitigating this vulnerability. This includes operators, platform administrators, vulnerability management teams, and security teams who need to verify exposure and take steps to mitigate or patch the vulnerability to prevent potential DoS attacks.

Why it matters

Defenders should prioritize verifying exposure of vLLM installations to this memory exhaustion vulnerability, especially for unauthenticated API access points, and take steps to mitigate or patch the vulnerability to prevent potential DoS attacks.

• Potential for host memory exhaustion leading to API process crashes
• Bypassing of --api-key authentication for /tokenize endpoint
• Need for verification of vLLM version and exposure
• Potential for denial-of-service (DoS) attacks

Technical summary

The vLLM project is vulnerable to a memory exhaustion attack in versions 0.25.0 before 0.31.0. An unauthenticated attacker can send distinct or oversized mel_config values to the /tokenize endpoint, bypassing the --api-key, to exhaust host memory and crash the API process. This vulnerability affects the MOSS-Audio model serving and can lead to potential denial-of-service (DoS) attacks. Defenders should prioritize verifying exposure of vLLM installations to this memory exhaustion vulnerability, especially for unauthenticated API access points.

Defensive priority

Defenders should prioritize verifying exposure of vLLM installations to this memory exhaustion vulnerability, especially for unauthenticated API access points.

Recommended defensive actions

• Verify vLLM version and apply patches if necessary
• Restrict access to the /tokenize endpoint
• Monitor API usage for suspicious activity
• Review compensating controls for exposed systems while remediation is scheduled and verified
• Check relevant monitoring, detection, and logs for exposed assets that need extra review
• Track exceptions, retest remediated assets, and close the item only after evidence is documented
• Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up

Evidence notes

The CVE record and source item detail a memory exhaustion vulnerability in vLLM versions 0.25.0 before 0.31.0, specifically in the MOSS-Audio model serving via the /tokenize endpoint. The vulnerability allows unauthenticated attackers to send distinct or oversized mel_config values to exhaust host memory and crash the API process. Defenders should verify exposure and consider patching or mitigating this vulnerability. Evidence is limited to CVE record and source item details.

Sources and references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.