Vulnerability · Published Sep 26, 2026 · Assigned by VulnCheck

CVE-2026-100647vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt

Medium 5.3

In plain English

A medium-severity improper input validation in vllm from vllm-project. There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.3% (higher than 29% of all known vulnerabilities).

Severity (CVSS 3.1)5.3Medium

Exploit odds (30 days)0.3%top 71.4% of all CVEs

Exploited in the wildNo reportsCISA SSVC: none

Attack toolingNoneno public tooling known

Official description

vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, which lacks maximum length validation and is processed on the single EngineCore scheduler thread. Unauthenticated attackers can send HTTP requests with multi-hundred-megabyte salt values that trigger expensive pickle serialization and SHA-256 hashing, stalling the scheduler thread and denying service to all concurrent requests.

Affected products (as reported)

Vendor	Product	Versions
vllm-project	vllm	0 – before 0.29.0

Weakness: CWE-20 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L

References

GitHub Security Advisory (GHSA-wpww-v874-ph2p) vendor-advisory
VulnCheck Advisory: vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt third-party-advisory