UTC
Critical

SGLang: Unauthenticated RCE via --enable-custom-logit-processor

CVE-2026-7304 · GHSA-36m8-w8qf-g76p · published 2026-05-18 · updated 2026-05-29

Severity
Critical
CVSS score
9.8
Chance of exploitation (EPSS)
0.58%
Fixed in
No fix yet
  • Exploitation No known exploitation
  • Automatable Yes
  • Technical impact Total
  • CISA SSVC
What it is

SGLang's multimodal generation runtime is vulnerable to unauthenticated remote code execution when the --enable-custom-logit-processor option is enabled, as Python objects loaded via dill.loads() will be deserialized without validation.

What to do

No fixed release is listed yet for sglang. Follow the source advisory for updates, and until one ships:

General guidance from ShieldPi, not a vendor fix

  • Command or code injection. Remove or disable the affected server or tool, or run it in a sandbox with no secrets in its environment.
  • Unsafe deserialization. Do not let it load files, models or data from sources you do not trust, and keep its RPC and socket ports reachable only from localhost or hosts you trust.
  • Missing or bypassable authentication. Do not expose it on a network others can reach, and put authentication you control in front of it, such as an authenticating proxy or a VPN. Binding it to localhost is not enough on its own: a web page open in your browser can still reach it, so require a token as well.
Affected
PackageRegistryAffectedFixed in
sglangPyPI>= 0.4.1.post7, <= 0.5.12No fix yet

The impact, patch and workaround text comes from the advisory, shortened (GitHub Advisory Database, CC BY 4.0). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.

Sources: GitHub Advisory Database, reviewed and malware advisories (CC BY 4.0, https://github.com/advisories); advisories published by Cursor (https://github.com/cursor/cursor/security/advisories); advisories published by OpenClaw (https://github.com/openclaw/openclaw/security/advisories); advisories published by vLLM (https://github.com/vllm-project/vllm/security/advisories); advisories published by Open WebUI (https://github.com/open-webui/open-webui/security/advisories); advisories published by n8n (https://github.com/n8n-io/n8n/security/advisories); advisories published by Flowise (https://github.com/FlowiseAI/Flowise/security/advisories); advisories published by Langflow (https://github.com/langflow-ai/langflow/security/advisories); advisories published by the MCP Python SDK project (https://github.com/modelcontextprotocol/python-sdk/security/advisories); advisories published by MLflow (https://github.com/mlflow/mlflow/security/advisories); CISA Known Exploited Vulnerabilities catalog (CC0). Data as of 2026-11-22. ShieldMap describes how each problem works and how to fix it; it does not publish exploit code. Report a correction to [email protected].