PYSEC-2026-2018
Vulnerability from pysec - Published: 2026-07-07 16:03 - Updated: 2026-07-07 17:25Summary
A memory corruption vulnerability that leading to a crash (denial-of-service) and potentially remote code execution (RCE) exists in vLLM versions 0.10.2 and later, in the Completions API endpoint. When processing user-supplied prompt embeddings, the endpoint loads serialized tensors using torch.load() without sufficient validation.
Due to a change introduced in PyTorch 2.8.0, sparse tensor integrity checks are disabled by default. As a result, maliciously crafted tensors can bypass internal bounds checks and trigger an out-of-bounds memory write during the call to to_dense(). This memory corruption can crash vLLM and potentially lead to code execution on the server hosting vLLM.
Details
A vulnerability that can lead to RCE from the completions API endpoint exists in vllm, where due to missing checks when loading user-provided tensors, an out-of-bounds write can be triggered. This happens because the default behavior of torch.load(tensor, weights_only=True) since pytorch 2.8.0 is to not perform validity checks for sparse tensors, and this needs to be enabled explicitly using the torch.sparse.check_sparse_tensor_invariants context manager.
The vulnerability is in the following code in vllm/entrypoints/renderer.py:148
def _load_and_validate_embed(embed: bytes) -> EngineEmbedsPrompt:
tensor = torch.load(
io.BytesIO(pybase64.b64decode(embed, validate=True)),
weights_only=True,
map_location=torch.device("cpu"),
)
assert isinstance(tensor, torch.Tensor) and tensor.dtype in (
torch.float32,
torch.bfloat16,
torch.float16,
)
tensor = tensor.to_dense()
Because of the missing checks, loading invalid prompt embedding tensors provided by the user can cause an out-of-bounds write in the call to to_dense .
Impact
All users with access to this API are able to exploit this vulnerability. Unsafe deserialization of untrusted input can be abused to achieve DoS and potentially remote code execution (RCE) in the vLLM server process. This impacts deployments running vLLM as a server or any instance that deserializes untrusted/model-provided payloads.
Fix
https://github.com/vllm-project/vllm/pull/27204
Acknowledgements
Finder: AXION Security Research Team (Omri Fainaro, Bary Levy): discovery and coordinated disclosure.
| Name | purl | vllm | pkg:pypi/vllm |
|---|
{
"affected": [
{
"package": {
"ecosystem": "PyPI",
"name": "vllm",
"purl": "pkg:pypi/vllm"
},
"ranges": [
{
"events": [
{
"introduced": "0.10.2"
},
{
"fixed": "0.11.1"
}
],
"type": "ECOSYSTEM"
}
],
"versions": [
"0.10.2",
"0.11.0"
]
}
],
"aliases": [
"CVE-2025-62164",
"GHSA-mrw7-hf4f-83pf"
],
"details": "### Summary\nA memory corruption vulnerability that leading to a crash (denial-of-service) and potentially remote code execution (RCE) exists in vLLM versions 0.10.2 and later, in the Completions API endpoint. When processing user-supplied prompt embeddings, the endpoint loads serialized tensors using torch.load() without sufficient validation.\n\nDue to a change introduced in PyTorch 2.8.0, sparse tensor integrity checks are disabled by default. As a result, maliciously crafted tensors can bypass internal bounds checks and trigger an out-of-bounds memory write during the call to to_dense(). This memory corruption can crash vLLM and potentially lead to code execution on the server hosting vLLM.\n\n### Details\nA vulnerability that can lead to RCE from the completions API endpoint exists in vllm, where due to missing checks when loading user-provided tensors, an out-of-bounds write can be triggered. This happens because the default behavior of `torch.load(tensor, weights_only=True)` since pytorch 2.8.0 is to not perform validity checks for sparse tensors, and this needs to be enabled explicitly using the [torch.sparse.check_sparse_tensor_invariants](https://docs.pytorch.org/docs/stable/generated/torch.sparse.check_sparse_tensor_invariants.html) context manager.\n\nThe vulnerability is in the following code in [vllm/entrypoints/renderer.py:148](https://github.com/vllm-project/vllm/blob/a332b84578cdc0706e040f6a765954c8a289904f/vllm/entrypoints/renderer.py#L148)\n\n```python\n def _load_and_validate_embed(embed: bytes) -\u003e EngineEmbedsPrompt:\n tensor = torch.load(\n io.BytesIO(pybase64.b64decode(embed, validate=True)),\n weights_only=True,\n map_location=torch.device(\"cpu\"),\n )\n assert isinstance(tensor, torch.Tensor) and tensor.dtype in (\n torch.float32,\n torch.bfloat16,\n torch.float16,\n )\n tensor = tensor.to_dense()\n```\n\nBecause of the missing checks, loading invalid prompt embedding tensors provided by the user can cause an out-of-bounds write in the call to `to_dense` .\n\n### Impact\nAll users with access to this API are able to exploit this vulnerability. Unsafe deserialization of untrusted input can be abused to achieve DoS and potentially remote code execution (RCE) in the vLLM server process. This impacts deployments running vLLM as a server or any instance that deserializes untrusted/model-provided payloads.\n\n## Fix\n\nhttps://github.com/vllm-project/vllm/pull/27204\n\n## Acknowledgements\n\nFinder: AXION Security Research Team (Omri Fainaro, Bary Levy): discovery and coordinated disclosure.",
"id": "PYSEC-2026-2018",
"modified": "2026-07-07T17:25:47.682656Z",
"published": "2026-07-07T16:03:10.617331Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-mrw7-hf4f-83pf"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2025-62164"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/pull/27204"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/commit/58fab50d82838d5014f4a14d991fdb9352c9c84b"
},
{
"type": "PACKAGE",
"url": "https://github.com/vllm-project/vllm"
},
{
"type": "PACKAGE",
"url": "https://pypi.org/project/vllm"
},
{
"type": "ADVISORY",
"url": "https://github.com/advisories/GHSA-mrw7-hf4f-83pf"
}
],
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
],
"summary": "vLLM deserialization vulnerability leading to DoS and potential RCE"
}
Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Vulnerability timeline
Sightings and recorded dates, grouped by day · UTC
How durations are calculated
Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.
Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.
Loading timeline…
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.