# CVE-2026-5760 — SGLang

> As of 2026-11-11, CVE-2026-5760 is a CRITICAL-severity vulnerability in SGLang, CVSS v3.1 9.8, EPSS 0.8% (58.5th percentile). Threadlinqs Intelligence links 1 tracked threat campaign to CVE-2026-5760, most recently “SGLang RCE Exploited Against NEAR AI Inference Server — Data Stolen From 18 Agents and ~100 Users”.

**Last updated:** 2026-11-11

## What is CVE-2026-5760?

SGLang's reranking endpoint (/v1/rerank) achieves Remote Code Execution (RCE) when a model file containing a malcious tokenizer.chat_template is loaded, as the Jinja2 chat templates are rendered using an unsandboxed jinja2.Environment().

The record classifies CVE-2026-5760 under weakness class [CWE-94](https://intel.threadlinqs.com/cwe/CWE-94). Its CVSS v3 base vector states that the flaw is reachable remotely over the network, needs no prior authentication, needs no user interaction, and has high impact on confidentiality, integrity, availability. 1 affected-product entry is recorded, across 1 vendor, listed below. The identifier was first published 205 days ago.

## Severity and exploitation probability

- **CVSS v3.1 base score:** 9.8 — CRITICAL (`CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H`)
- **EPSS (FIRST):** 0.8% probability of exploitation in the next 30 days, higher than 58.5% of all scored CVEs
- **CISA KEV:** Not listed in the CISA Known Exploited Vulnerabilities catalog
- **Threadlinqs priority:** 9.2/10 — a Threadlinqs composite of the CVSS base score, the EPSS percentile and public exploit availability
- **Published:** 2026-04-20

## Is CVE-2026-5760 being exploited?

Weaponised exploit code for CVE-2026-5760 is publicly available. 2 public proof-of-concept repositories are tracked for this identifier. It currently carries a trending score of 28 in the Threadlinqs vulnerability feed.

- [Stuub/SGLang-0.5.9-RCE](https://github.com/Stuub/SGLang-0.5.9-RCE) (github)
- [trickest/cve](https://github.com/trickest/cve/blob/main/2026/CVE-2026-5760.md) (trickest)

## Affected products and versions

- **SGLang**: SGLang

## How to fix CVE-2026-5760

The record marks a vendor fix as available for CVE-2026-5760. Patch reference: [https://github.com/sgl-project/sglang/pull/23660](https://github.com/sgl-project/sglang/pull/23660). Apply the vendor fix referenced above to every affected product listed in this record, then confirm the running version against the vendor advisory.

## Threat activity tracking CVE-2026-5760

1 tracked threat in the Threadlinqs corpus references CVE-2026-5760, either in the campaign’s CVE list or as an indicator on the campaign record.

- [SGLang RCE Exploited Against NEAR AI Inference Server — Data Stolen From 18 Agents and ~100 Users](https://intel.threadlinqs.com/threat/TL-2026-5697) — HIGH · 2026-11-11

## Sources

Enriched from CVE.org, NVD, FIRST EPSS, GitHub Security Advisories, public proof-of-concept repositories. This product uses the NVD API but is not endorsed or certified by the NVD.

**Other references**

- [github.com](https://github.com/Stuub/SGLang-0.5.9-RCE)
- [github.com (23660)](https://github.com/sgl-project/sglang/pull/23660)

Canonical: https://intel.threadlinqs.com/cve/CVE-2026-5760
Every threat comes with one complete detection in Splunk SPL, Microsoft KQL and Sigma, free for everyone; a free account adds daily TL Creds to unlock the rest. Sign up free (Blue): https://threadlinqs.com/?modal=signup
Indicator values unlock with TL Creds (https://threadlinqs.com/tl-creds); MCP access for AI agents needs Purple: https://intel.threadlinqs.com/mcp