Technology · SGLang Project
SGLang Project SGLang vulnerabilities
Updated . Rebuilt every hour.
Junglewise Threat Intelligence has tracked 26 vulnerabilities in SGLang Project SGLang: 0 in the last 7 days and 12 in the last 90 days, 5 of them critical and 0 exploited in the wild. The most recent, CVE-2026-112270, was published on 19 October 2026. 8 vulnerabilities in the last four weeks, up 167% on the four before (3).
- Last 7 days
- 0
- Last 90 days
- 12
- Critical, all time
- 5
- Exploited in the wild
- 0
About SGLang Project SGLang
SGLang is a structured generation language designed for large language models, providing a fast serving framework and a domain-specific language for programming LLMs.
Latest SGLang Project SGLang vulnerabilities
- CVE-2026-112270: SGLang uncontrolled memory allocation in Mooncake disaggregation backendmediumCVSS 6.5
- CVE-2026-112269: SGLang Mooncake disaggregation backend unauthenticated denial of servicemediumCVSS 6.5
- CVE-2026-112267: SGLang server-side request forgery in prefill/decode disaggregationmediumCVSS 6.5
- CVE-2026-111392: SGLang cache poisoning via untrusted mm_hashes in multimodal prefix cachemediumCVSS 4.8
- CVE-2026-111391: SGLang code injection in _evaluate_filter()highCVSS 8.1
- CVE-2026-109229: SGLang denial of service via disabled image decompression protectionhighCVSS 7.5EPSS 0.4%
- CVE-2026-109228: SGLang resource budget bypass in GLM-V video processinghighCVSS 7.5EPSS 0.5%
- CVE-2026-109226: SGLang decompression amplification in load_audiohighCVSS 7.5EPSS 0.5%
- CVE-2026-93838: SGLang unbounded memory allocation in handle_staging_reqmediumCVSS 5.9EPSS 0.7%
- CVE-2026-93688: SGLang denial of service in prefill/decode disaggregation mode with Mooncake backendhighCVSS 7.5EPSS 0.7%
- CVE-2026-92972: SGLang unauthenticated PUT /route endpoint KV transfer poisoninghighCVSS 8.6EPSS 0.5%
- CVE-2026-86793: SGLang unsafe pickle deserialization in /update_weights_from_tensorcriticalCVSS 9.8EPSS 0.8%
- CVE-2026-15978: SGLang model weight exfiltration via untrusted distributed peer enrollmentinfoCVSS 7.5
- CVE-2026-15977: SGLang credential leakage in server_info endpointinfoCVSS 7.5
- CVE-2026-15976: SGLang insecure pickle deserialization in weight update endpointinfoCVSS 9.8
- CVE-2026-15974: SGLang SSRF and local file read in multimodal chat endpointinfoCVSS 7.5
- CVE-2026-15971: SGLang RCE via eval sandbox escape in dumper subsysteminfoCVSS 9.8
- CVE-2026-15969: SGLang SafeUnpickler bypass RCE in load_lora_adapter_from_tensorsinfoCVSS 9.8
- CVE-2026-14890: SGLang remote code execution in expert-parallel backup subsysteminfoCVSS 9.8
- CVE-2026-10775: sgl-project SGLang denial of service in Cache HandlerlowCVSS 3.6EPSS 0.1%
- CVE-2026-10300: SGLang reachable assertion in LoRA manager scheduler admissionlowCVSS 3.7EPSS 0.4%
- CVE-2026-7304: SGLang RCE via unsafe deserialization in custom logit processorcriticalCVSS 9.8EPSS 0.9%
- CVE-2026-7302: SGLang path traversal in multimodal generation runtimecriticalCVSS 9.1EPSS 0.6%
- CVE-2026-7301: SGLang RCE via insecure pickle deserialization in scheduler ROUTER socketcriticalCVSS 9.8EPSS 0.6%
- CVE-2026-5760: LMSYS SGLang RCE via unsandboxed Jinja2 template in rerank endpointcriticalCVSS 9.8EPSS 0.4%
Most severe SGLang Project SGLang vulnerabilities
Exploited in the wild first, then by severity and CVSS score.
- CVE-2026-7304: SGLang RCE via unsafe deserialization in custom logit processorcriticalCVSS 9.8EPSS 0.9%
- CVE-2026-86793: SGLang unsafe pickle deserialization in /update_weights_from_tensorcriticalCVSS 9.8EPSS 0.8%
- CVE-2026-7301: SGLang RCE via insecure pickle deserialization in scheduler ROUTER socketcriticalCVSS 9.8EPSS 0.6%
- CVE-2026-5760: LMSYS SGLang RCE via unsandboxed Jinja2 template in rerank endpointcriticalCVSS 9.8EPSS 0.4%
- CVE-2026-7302: SGLang path traversal in multimodal generation runtimecriticalCVSS 9.1EPSS 0.6%
- CVE-2026-92972: SGLang unauthenticated PUT /route endpoint KV transfer poisoninghighCVSS 8.6EPSS 0.5%
- CVE-2026-111391: SGLang code injection in _evaluate_filter()highCVSS 8.1
- CVE-2026-3989: SGLang unsafe pickle deserialization in replay_request_dump.pyhighCVSS 7.8EPSS 0.4%
- CVE-2026-93688: SGLang denial of service in prefill/decode disaggregation mode with Mooncake backendhighCVSS 7.5EPSS 0.7%
- CVE-2026-109228: SGLang resource budget bypass in GLM-V video processinghighCVSS 7.5EPSS 0.5%
Vulnerabilities per week
8 vulnerabilities in the last four weeks, up 167% on the four before (3). The last 13 weeks, by the week each vulnerability was published.
| Week of | Bar | Vulns | Critical |
|---|---|---|---|
| 17 Aug 2026 | 0 | 0 | |
| 24 Aug 2026 | 0 | 0 | |
| 31 Aug 2026 | 0 | 0 | |
| 7 Sep 2026 | 1 | 1 | |
| 14 Sep 2026 | 3 | 0 | |
| 21 Sep 2026 | 0 | 0 | |
| 28 Sep 2026 | 0 | 0 | |
| 5 Oct 2026 | 0 | 0 | |
| 12 Oct 2026 | 5 | 0 | |
| 19 Oct 2026 | 3 | 0 | |
| 26 Oct 2026 | 0 | 0 | |
| 2 Nov 2026 | 0 | 0 | |
| 9 Nov 2026 | 0 | 0 |
Embed a badge
A live badge for a README, a status page or an internal wiki: how many SGLang Project SGLang vulnerabilities were published in the last 7 days and how many of them are critical, refreshed hourly.
[](https://junglewise.ai/threats/technologies/sglang)<a href="https://junglewise.ai/threats/technologies/sglang"><img src="https://junglewise.ai/threats/technologies/sglang/badge.svg" alt="SGLang Project SGLang vulnerabilities" /></a>How this is built
Junglewise Threat Intelligence collects vulnerabilities from NVD, GitHub Security Advisories, OSV, the CISA Known Exploited Vulnerabilities catalog, FIRST EPSS and vendor advisories, and matches each one to the technologies and vendors it affects. Dates are the date a vulnerability was published, in UTC.
The pages are rebuilt from the database every hour. Frozen weekly and monthly reports never change once published, so they can be cited.
Use this data
The same data is at https://junglewise.ai/threats/technologies/sglang.json, and this page as text at https://junglewise.ai/threats/technologies/sglang.md, for scripts and language models. It is free to reuse under CC BY 4.0 with a link back to this page.
Cite as: Junglewise Threat Intelligence, "SGLang Project SGLang vulnerabilities", https://junglewise.ai/threats/technologies/sglang, 12 November 2026.