En bref
- CVE-2026-5497 est une vulnérabilité ÉLEVÉE (CVSS 7.5) touchant vllm, vllm (vllm).
- Probabilité d'exploitation dans les 30 prochains jours (EPSS) : 0.54%, plus élevée que 45.2% des vulnérabilités.
- Correctif : mettre à jour vers 0.19.0 ou plus récent.
| Package | Affected versions | Patched versions |
|---|
| vllm PyPI | ≥ 0.8.0, < 0.19.0
| 0.19.0
|
| vllm | ≥ 0.8.0, < 0.19.0
| 0.19.0
|
Description
vLLM versions 0.8.0 and later are vulnerable to an Out-of-Memory (OOM) Denial of Service (DoS) attack due to unbounded frame count processing in the VideoMediaIO.load_base64() method. When processing video/jpeg data URLs, the method splits the base64 data string on commas to extract individual JPEG frames without enforcing a frame count limit. An attacker can exploit this by crafting a single API request containing thousands of comma-separated base64-encoded JPEG frames in a data URL, causing the server to decode all frames into memory and crash due to excessive memory consumption. This vulnerability is reachable via the OpenAI-compatible chat completions API and does not require authentication.
NVD description
vLLM versions 0.8.0 and later are vulnerable to an Out-of-Memory (OOM) Denial of Service (DoS) attack due to unbounded frame count processing in the `VideoMediaIO.load_base64()` method. When processing `video/jpeg` data URLs, the method splits the base64 data string on commas to extract individual JPEG frames without enforcing a frame count limit. An attacker can exploit this by crafting a single API request containing thousands of comma-separated base64-encoded JPEG frames in a data URL, causing the server to decode all frames into memory and crash due to excessive memory consumption. This vulnerability is reachable via the OpenAI-compatible chat completions API and does not require authentication.
Running one of these products? Scan your site to see whether it exposes a vulnerable version.
Scan my site free