vLLM: Request-selected PyNvVideoCodec GPU decode bypasses static VRAM reservation
Noroxi assessment
Priority: Monitor (action score 26/100)
- The sources we track (CISA KEV, Exploit-DB, Metasploit, Nuclei, GitHub) show no public exploit code or exploitation in the wild.
- Probability of exploitation in the next 30 days: 0.58% (FIRST EPSS).
- Anyone who can reach it over the internet can trigger it. A low-privileged account is enough. No user action is required. No special conditions are required; it is repeatable.
- Fixed version: vllm (PyPI) 0.28.0.
Low priority for now. Upgrade to vllm (PyPI) 0.28.0 or later.
Generated automatically from the record's data: CVSS (NVD and CNA), CISA KEV and SSVC, FIRST EPSS, exploit sources, fixed versions. How we calculate it
vLLM is an inference and serving engine for large language models. Prior to 0.28.0, request bodies for Chat Completions and Responses can set media_io_kwargs.video.video_backend to pynvvideocodec, and MediaConnector.fetch_video forwards that choice to VideoMediaIO even when startup configuration selected a software decoder. The engine's _reserve_mm_ipc_gpu_memory logic budgets decoder memory only from static configuration, so the request-selected VIDEO_LOADER_REGISTRY backend can create a CUDA context, decoder surfaces, and decoded-frame allocations that were not removed from the engine's KV-cache budget. An attacker able to submit video requests to a video-capable GPU deployment with PyNvVideoCodec installed can exhaust shared GPU memory, causing request failures, worker crashes, or denial of service. The first release containing the fix is version 0.28.0.
- Published
- Sep 16, 2026
- Updated
- Oct 7, 2026
- EPSS
- 0.6% · 46th percentile
- CWE
- CWE-400
Report tools
Action score
26
Monitor
Low priority for now.
- CVSS
- 26 / 40 · 6.5 / 10
- CISA KEV
- 0 / 30 · Not listed
- EPSS
- 0 / 30 · 0.6%
CISA SSVC decision
- Exploitation
- proof of concept
- Automatable
- no
- Technical impact
- partial
Vulnrichment: CISA's decision-tree inputs.
CNA vs NVD score
- NVD
- —
- CNA · GitHub_M
- 6.5
- NVD has not scored this yet; the score shown is the CNA’s.
The score given by the assigning authority versus NVD’s independent score. A gap means the severity is contested.
Get told when this record changes
- Not in KEV: no record of exploitation in the wild
- No public exploit code on record
- A fix has been released
If any of these changes, we'll tell you in the daily email digest.
Running WordPress? Scan your site free with WP Lens
Noroxi analysis
What does it mean in your environment?
The assessment above is generated automatically from the record's data; it doesn't know your environment. Our team verifies on your systems whether this vulnerability is actually exploitable and which mitigation is enough.
We use this product, ask for helpAffected systems
Is your version affected?
Pick the product and type the version you run. It's compared with this record's version ranges in your browser; the version isn't sent anywhere.
| Vendor | Product | CPE |
|---|---|---|
| vllm | vllm | cpe:2.3:a:vllm:vllm |
Affected versions
NVD version ranges (for catalog products). Add the product to your stack with a version and matching uses these.
- vllm vllmbefore 0.28.0
Versions reported by the vendor
Affected version ranges reported by the assigning authority (GitHub_M). Independent of NVD's CPE analysis and usually ahead of it.
vllm-project vllm
- < 0.28.0affected
Package-level exposure
OSV and GitHub Advisory data: ecosystem, package and range. SBOM matching uses this table.
| Ecosystem | Package | Affected range | Fix |
|---|---|---|---|
| PyPI | vllm | before 0.28.0 | 0.28.0 |
Same product
vllm: all recordsOther highest-scoring records for the same primary product.
- CVE-2026-22778vLLM leaks a heap address when PIL throws an error44Plan
- CVE-2025-32444vLLM Vulnerable to Remote Code Execution via Mooncake Integration40Plan
- CVE-2026-25960SSRF Protection Bypass in vLLM39Monitor
- CVE-2026-22807vLLM affected by RCE via auto_map dynamic module loading during model initialization39Monitor
- CVE-2025-47277vLLM Allows Remote Code Execution via PyNcclPipe Communication Service39Monitor
- CVE-2024-11041Remote Code Execution in vllm-project/vllm39Monitor
Remediation
Which version to upgrade to
Fix versions compiled from the vendor, package registries and Microsoft. Verify the vendor's note before upgrading.
| Product / package | Fixed version | Source |
|---|---|---|
| PyPI:vllm | 0.28.0 | Package registry (OSV) |
Exploit status
No known public exploit
No public exploit has been observed yet. That doesn't mean you're safe, only that the bar is a little higher.
Research context
For pentesters and researchers: attack profile, score disagreement, timeline, patch commits, credits, variant and chain candidates, bug bounty scope. All derived from existing data; no exploit code.
Timeline
From publication to today: proof of concept, Metasploit module, CISA KEV and fix record. Dates are as reported by the sources.
No dated events beyond publication.
FIRST EPSS daily score; only changes of 0.01 or more are recorded (step chart).
Patch and commit links
Commit, PR and diff links among the references. A starting point for patch-diffing and variant hunting; fixes, not exploits.
Credits
All researchersFinders, reporters and analysts named in the CNA record. Click a name for that researcher’s other records.
No credits in the CNA record.
Variant candidates
Same product, same weakness class, within 18 months. If the patch missed the root cause, the sibling bug is here.
- CVE-2026-3723719 days apartvLLM up to and including 0.17.0 allows remote attackers to cause a Denial of Service via memory exhaustion.30Monitor
- CVE-2026-549797 days apartUnbounded Frame Count in video/jpeg Base64 Data URL Processing Leads to OOM DoS in vllm-project/vllm30Monitor
- CVE-2025-48956391 days apartvLLM API endpoints vulnerable to Denial of Service Attacks30Monitor
- CVE-2026-10065010 days apartvLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization28Monitor
- CVE-2026-10065110 days apartvllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass28Monitor
- CVE-2026-905544 days apartvLLM before 0.28.0 Denial of Service via audio extraction27Monitor
Chain candidates
An authentication bypass and a privilege-requiring bug in the same product, published close together: combined they may become an unauthenticated path.
—
Bug bounty scope
No known public program.
Source: bounty-targets-data (public HackerOne, Bugcrowd, Intigriti, YesWeHack listings).
Technical details
Attack conditions
- Anyone who can reach it over the internet can trigger it.
- A low-privileged account is enough.
- No user action is required.
- No special conditions are required; it is repeatable.
If successful
- Confidentiality
- none
- Integrity
- none
- Availability
- high · the service can be disrupted
- Attack vector
- Network
- Attack complexity
- Low
- Privileges required
- Low
- User interaction
- None
- Scope
- Unchanged
- Confidentiality impact
- None
- Integrity impact
- None
- Availability impact
- High
Weakness class (CWE)
CWE-400 · Uncontrolled Resource ConsumptionCVSS vector
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
nvd-secondary
Root cause
Code that parses nested structures sets no depth or size limit. A small input can turn into a disproportionate processing load.
The bug in code
The whole CWE-400 classA representative example of this vulnerability class. Not the vendor's source code; it shows the faulty pattern and its fix.
Vulnerable
function expand(node) { return node.children.map(expand);}Fixed
const MAX_DEPTH = 32;function expand(node, depth = 0) { if (depth > MAX_DEPTH) throw new Error("too deep"); return node.children.map((c) => expand(c, depth + 1));}Attack context
MITRE CAPEC attack patterns and ATT&CK techniques for this weakness class (CWE). A starting point for detection rules and threat hunting.
Change log
- Fix✗ → ✓
For records you follow, these changes also arrive as notifications. →
References
Show all references (2 more)Show less
Vendor advisories and official records. Exploit/PoC links are deliberately left out.