CVE-2026-93592: Input validation flaw in vLLM embeddings and pooling endpoints
vLLM
vLLM versions before 0.28.0 fail to validate lower bounds of token IDs in embedding endpoints, enabling unauthenticated attackers to crash the engine. Submitting negative token IDs triggers a CUDA device assertion that poisons the GPU context.
disclosed CVENo confirmed exploitation, so this record is tracked as a disclosed vulnerability and not counted as an incident. It moves into the counts if CISA or the vendor confirms exploitation. Why
- CVSS
- 8.7
- EPSS
- 0.01 · top 54% · 2026-11-13
- fix status
- Fix status unknown
- first report
- 18 Sep 2026
- last update
- 18 Sep 2026
- CVE ids
Timeline · 1
Sources · 1
Advisories and databases
- 2026-09-18CVE-2026-93592: vLLM versions before 0.28.0 fail to validate the lower bound of token IDs in the /v1/embeddings and /pooling endpoints, allowing…NVD · nvd.nist.gov/vuln/detail/CVE-2026-93592
Titles and links belong to their publishers. The summary above is written by gcdRadar in its own words; data is CC BY-SA 4.0.
Classification
- lane
- Patch now
- AI role
- Attack on AI
- categories
- Infrastructure vulnerability
- severity
- 2 · Disclosed vulnerability
- fix status
- Fix status unknown
- targets
- Framework / tooling
- owasp llm top 10
- LLM03:2025 Supply Chain
- mitre atlas
- no technique for these categories
- dates
- happened 18 Sep 2026 · first reported 18 Sep 2026
- origin
- NVDVulnerability record from NVD / the GitHub Advisory Database. This product uses the NVD API but is not endorsed or certified by the NVD.
- record
- confidence 0.95 · created 2026-10-18 06:33Z · updated 2026-10-18 06:33Z
Lane, AI role, categories and severity come from the classifier; the OWASP and ATLAS rows are derived from the categories. How these are assigned
Copy a brief
This record as a short Markdown brief, for a ticket, a chat or a coding agent.
the brief as text
# CVE-2026-93592: Input validation flaw in vLLM embeddings and pooling endpoints https://gcdradar.vercel.app/incidents/cve-2026-93592-input-validation-flaw-in-vllm-embeddings-and · AIR-2026-0150 · closed 2026-09-18 - Lane: Patch now - Severity: 2 (Disclosed vulnerability) - Fix status: Fix status unknown - First reported: 2026-09-18 - AI role: Attack on AI - Products: vLLM - CVE ids: CVE-2026-93592 - CVSS: 8.7 ## Summary vLLM versions before 0.28.0 fail to validate lower bounds of token IDs in embedding endpoints, enabling unauthenticated attackers to crash the engine. Submitting negative token IDs triggers a CUDA device assertion that poisons the GPU context. ## What to do No action recorded; read the sources below and the vendor's advisory. ## Latest updates - 2026-09-18: CVE-2026-93592: vLLM versions before 0.28.0 fail to validate the lower bound of token IDs in the /v1/embeddings and /pooling endpoints, allowing… (https://nvd.nist.gov/vuln/detail/CVE-2026-93592) ## Sources - 2026-09-18 NVD: CVE-2026-93592: vLLM versions before 0.28.0 fail to validate the lower bound of token IDs in the /v1/embeddings and /pooling endpoints, allowing… https://nvd.nist.gov/vuln/detail/CVE-2026-93592 Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.
Related
- closed7.1vLLM CVE-2026-109311: vLLM audio transcription endpoint heap overflow vulnerability
Patch now · Disclosed · since 12 Oct 2026 · CVE-2026-109311
AI - closed4.3vLLM CVE-2026-105775: Out-of-bounds read vulnerability in vLLM completions request handler
Patch now · Disclosed · since 6 Oct 2026 · CVE-2026-105775
AI - closed4.3vLLM Denial of Service Vulnerability in vLLM Penalty Handler
Patch now · Disclosed · since 6 Oct 2026 · CVE-2026-105922
AI - closed5.3vLLM vLLM Denial of Service via Unbounded Video Sampler Frames
Patch now · Disclosed · since 5 Oct 2026 · CVE-2026-105758
AI - closed3.1vLLM vLLM tool continuations drop cache_salt to restore prefix-cache oracle
Patch now · Disclosed · since 5 Oct 2026 · updated 6 Oct 2026 · CVE-2026-105752
AI - closed5.3vLLM vLLM GLMGA video sampling permits CPU and memory exhaustion
Patch now · Disclosed · since 5 Oct 2026 · CVE-2026-105760
AI