CVE-2026-100650: DoS (CVSS 6.5)

Description

vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths — the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route — the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact.

Is your site exposed to CVE-2026-100650?

Run a free security scan — no signup, results in seconds.

CVSS v3.1 Score

6.5

MEDIUM

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

EPSS — Exploit Prediction

0.0065

Probability of exploitation

0.50%

Percentile rank

EPSS estimates the probability that this vulnerability will be exploited in the wild within the next 30 days. A higher score means more likely to be exploited.

Weakness Type (CWE)

CWE-400 Uncontrolled Resource Consumption

References

Other References

https://github.com/vllm-project/vllm/commit/752a3a504485790a2e8491cacbb35c137339ad34 https://github.com/vllm-project/vllm/security/advisories/GHSA-p6g9-7v3x-m8mv https://www.vulncheck.com/advisories/vllm-before-0.29.0-resource-exhaustion-via-unbounded-media-materialization https://github.com/vllm-project/vllm/security/advisories/GHSA-p6g9-7v3x-m8mv

Frequently Asked Questions

What is CVE-2026-100650? +

vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths — the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route — the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact. It has a CVSS v3.1 base score of 6.5 (MEDIUM).

How severe is CVE-2026-100650? +

CVE-2026-100650 has a CVSS v3.1 score of 6.5 out of 10, rated MEDIUM. This is a medium-severity vulnerability that should be remediated as part of regular maintenance. The EPSS score is 0.0065, placing it in the 0th percentile for exploitation probability.

How do I check if I'm vulnerable to CVE-2026-100650? +

You can use Secably's free Website Scanner to check your website for known vulnerabilities. For infrastructure scanning, use the Port Scanner to identify exposed services that may be affected. Check the vendor advisories linked above for specific patch and version information.

Related Vulnerabilities

CVE-2026-45802

FPDI is a collection of PHP classes that facilitate reading pages from existing PDF documents and using them as templates …

CVE-2026-67318

axios versions >=1.13.0 (Node.js HTTP adapter) fail to enforce the configured maxBodyLength limit on streamed request bodies when requests are …

CVE-2026-42343

FastGPT is an AI Agent building platform. In versions 4.14.13 and prior, the code-sandbox component suffers from insufficient resource isolation …

CVE-2025-41361

Uncontrolled resource consumption vulnerability in IDF v0.10.0-0C03-03 and ZLF v0.10.0-0C03-04. The devices improperly handle TLS requests associated with PROCOME sockets, …

CVE-2025-41360

Uncontrolled resource consumption vulnerability in IDF v0.10.0-0C03-03 and ZLF v0.10.0-0C03-04. The device is vulnerable to a packet flooding denial of …

CVE-2024-5422

An uncontrolled resource consumption of file descriptors in SEH Computertechnik utnserver Pro, SEH Computertechnik utnserver ProMAX, SEH Computertechnik INU-100 allows …

CVE-2026-100650