Servers and applications
CVE-2026-93592
Denial of service: a negative token id in `/v1/embeddings` or `/pooling` input kills the vLLM engine via a CUDA device-side assertion
- Severity
- High 8.7
- Published
- 3 Sept 2026
- Last changed
- 30 Sept 2026
- Added to UpHeron
- 3 Nov 2026
- Fixed in
- 0.28.0
- Exploit chance
- 0.5% in the next 30 daysHigher than 46% of CVEs. EPSS by FIRST, 19 Nov 2026
- Weakness
- CWE-129
- Attack
- Over the network · No login needed · No user action needed
- Attacker can
- Take it down
- CVSS vector
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X- Also known as
- GHSA-25q3-v2hm-8vpf
- Sources
- NVD, Project advisories (GitHub)
Description
A single request carrying a negative token id in the token-id form of the input field of /v1/embeddings or /pooling terminates the vLLM engine. The request is answered HTTP 500 and the server then serves nothing further: /health refuses the connection and every subsequent request from every client fails until the process is restarted.
No authentication, no prior request, and no unusual value are required. -1 is enough.
/v1/embeddings and /pooling accept token ids directly from the caller (input: list[int] | list[list[int]] | str | list[str] on CompletionRequestMixin, in vllm/entrypoints/pooling/base/protocol.py). The engine validated only the upper bound of those ids against the vocabulary, in InputProcessor._validate_model_input:
There was no corresponding lower-bound check, so a negative id passed validation and was used directly as an index into the embedding table on the GPU. That trips a device-side assertion inside PyTorch's index kernel:
Workarounds
For deployments that cannot upgrade to 0.28.0:
- Reject request bodies containing negative integers in input at a reverse proxy or API gateway in front of vLLM.
- Do not expose the token-id form of the input; accept only string input from untrusted callers.
- Restrict network access to /v1/embeddings and /pooling.
Process supervision is not a workaround: the request can simply be repeated.
Affected software
| Software | Affected versions |
|---|---|
vLLMServers and applications · cpe:a:vllm:vllm |
|
vLLMServers and applications · github:vllm-project/vllm |
|
About vLLM
open-source software for large language model inference
- Docker Hub pulls
- 44,195,225Docker Hub, 15 Nov 2026
- GitHub stars
- 96,943GitHub, 15 Nov 2026
- Forks
- 24,836GitHub, 15 Nov 2026
- Latest release
- v0.33.0rc3GitHub, 15 Nov 2026
- Released
- 2 Nov 2026GitHub
- License
- Apache Software License 2.0Wikidata
- First published on GitHub
- 9 Feb 2023GitHub
Website Source code Docker image
Vulnerabilities in vLLM
- In our list
- 169
- Published in the last 12 months
- 150
- Critical, last 12 months
- 6
- High, last 12 months
- 63
- Known exploited
- 0
- CVE-2026-130686vLLM 0.6.0 before 0.33.0 Resource Exhaustion via Multi-Image RequestsHigh 8.7 18 Nov 2026
- CVE-2026-130684vLLM before 0.33.0 Path Traversal via speculators_config verifier.name_or_pathMedium 5.3 18 Nov 2026
- CVE-2026-130685vLLM 0.28.0 before 0.33.0 Unauthenticated DoS via Out-of-Range Prompt Token IDHigh 8.7 18 Nov 2026
- GHSA-xgh8-qwjh-q853/v1/systemone bypasses the --trust-request-chat-template gateMedium 6.5 18 Nov 2026
- CVE-2026-129356vllm-project vllm Common Prefix single_type_kv_cache_manager.py race conditionLow 2.3 15 Nov 2026
- CVE-2026-126394vLLM 0.30.0 before 0.31.0 Resource Exhaustion via OpenPangu Video Backend num_framesMedium 6.9 13 Nov 2026
References
- NVD https://nvd.nist.gov/vuln/detail/CVE-2026-93592
- CVE.org https://www.cve.org/CVERecord?id=CVE-2026-93592
- GitHub Advisory Database https://github.com/advisories/GHSA-25q3-v2hm-8vpf
- github.com https://github.com/vllm-project/vllm/security/advisories/GHSA-25q3-v2hm-8vpf