GHSA-pmqf-x6x8-p7qw is a medium-severity (CVSS 6.5) CWE-129 vulnerability in vllm. A fix is available for vllm — see the affected versions and patch details below.
vLLM vulnerable to DoS with incorrect shape of multimodal embedding inputs
Exploitation Status
No confirmed exploitation observed yet
- CISA’s own triage has not observed active exploitation or public proof-of-concept code for this CVE as of its last assessment.
Exploitation and automatability from CISA’s SSVC triage for GHSA-pmqf-x6x8-p7qw.
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
How urgent is this, really
GHSA-pmqf-x6x8-p7qw plotted by exploitation likelihood (EPSS) against impact (CVSS). The shaded corner — EPSS 50%+ and CVSS 7.0+ — is where this CVE doesn't sit, though severity or exploitability alone can still warrant action.
Where this sits among everything scored
Of 377,166 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Real counts from FIRST.org, not a sample — log-scaled since the landscape is heavily right-skewed.
Real-World Exposure
vllmReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects PyPI packages — download data is not available via public APIs for these ecosystems.
Description
Summary
Users can crash the vLLM engine serving multimodal models by passing multimodal embedding inputs with correct ndim but incorrect shape (e.g. hidden dimension is wrong), regardless of whether the model is intended to support such inputs (as defined in the Supported Models page).
The issue has existed ever since we added support for image embedding inputs, i.e. #6613 (released in v0.5.5)
Details
Using image embeddings as an example:
- For models that support image embedding inputs, the engine crashes when scattering the embeddings to
inputs_embeds(mismatched shape) - For models that don't support image embedding inputs, the engine crashes when validating the inputs inside
get_input_embeddings(validation fails).
This happens because we only validate ndim of the tensor, but not the full shape, in input processor (via MultiModalDataParser).
Impact
- Denial of service by crashing the engine
Mitigation
- Use API key to limit access to trusted users.
- Set
--limit-mm-per-promptto 0 for all non-text modalities to ban multimodal inputs, which includes multimodal embedding inputs. However, the model would then only accept text, defeating the purpose of using a multi-modal model.
Resolution
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐍PyPI | vllm | ≥ 0.5.5&&< 0.11.1 | 0.11.1pip install --upgrade 'vllm==0.11.1' |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for vllm, including transitive dependencies — a direct dependency you never call can still pull in a vulnerable version.
Fix
Update vllm to 0.11.1 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms GHSA-pmqf-x6x8-p7qw is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 Security's impact-aware SCA analyses which vulnerable code paths your application actually calls, so a match like GHSA-pmqf-x6x8-p7qw can be triaged on real exposure rather than presence alone.
Tailored to GHSA-pmqf-x6x8-p7qw. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Fixing This On Your OS
If you run this on a Linux distribution, patch through your package manager against the distro's own security advisory below — it tracks the exact backported fix for your release, which can ship on a different timeline (and sometimes a different severity) than the upstream project.
This flaw is rated Moderate rather than Important because its impact is strictly limited to availability and requires low but existing privileges to exploit. The issue arises from incomplete shape validation of multimodal embedding tensors, which can cause deterministic crashes in the inference engine, but it does not…
| Product | Fixed in | Advisory |
|---|---|---|
| Red Hat AI Inference Server 3.2 | rhaiis/vllm-cuda-rhel9:3.2.5-1765552580 | RHSA-2025:23204 |
| Red Hat AI Inference Server 3.2 | rhaiis/vllm-rocm-rhel9:3.2.5-1765361180 | RHSA-2025:23205 |
| Red Hat AI Inference Server 3.2 | rhaiis/vllm-tpu-rhel9:3.2.5-1765552619 | RHSA-2025:23209 |
| Red Hat AI Inference Server 3.2 | rhaiis/vllm-rocm-rhel9:3.2.5-1765552603 | RHSA-2025:23449 |
| Red Hat AI Inference Server 3.2 | rhaiis/vllm-cuda-rhel9:1772160593 | RHSA-2026:3461 |
| Red Hat AI Inference Server 3.2 | rhaiis/vllm-rocm-rhel9:1772160625 | RHSA-2026:3462 |
Frequently Asked Questions
Is GHSA-pmqf-x6x8-p7qw in your dependencies?
O3 Security finds GHSA-pmqf-x6x8-p7qw across PyPI dependencies, including transitive ones, and its impact-aware SCA ranks findings by whether your code actually calls the vulnerable path.