GHSA-542M-9234-8GVF
Vulnerability from github – Published: 2026-10-28 21:14 – Updated: 2026-10-28 21:14Summary
vLLM renders chat templates by passing the resolved Jinja2 source to tokenizer.apply_chat_template(), which compiles and .render()s it in an ImmutableSandboxedEnvironment. The sandbox prevents code-execution escapes but imposes no wall-clock timeout, no output-size cap, and no nested-loop or binop limit. The render runs in a ThreadPoolExecutor whose max_workers defaults to 1, and CPython worker threads cannot be interrupted. An authenticated API client (or a malicious model repo loaded with trust_remote_code=False) can therefore submit a single short template and permanently wedge all chat-style endpoints (/v1/chat/completions, /v1/responses, /tokenize, /pooling-chat, /invocations-chat) for every tenant until the process is restarted. A string-multiplication variant instead OOM-kills the API-server process. Because of a separately-reported (GHSA-jmjw-q4x3-hmc2) gate omission on /v1/responses, the remote authenticated attacker vector works in default vllm serve configuration with no operator opt-in.
Impact
A single authenticated request of <100 bytes permanently denies all chat-style inference (/v1/chat/completions, /v1/responses, /tokenize, /pooling-chat, /invocations-chat) to every tenant of the API-server replica until the operator restarts the process. The string-multiplication variant OOM-kills the process. In a multi-tenant deployment this is a cross-tenant denial-of-service from any one tenant. From the malicious-model angle, a model repo whose tokenizer_config.json carries the hostile template wedges every replica that loads it on the first legitimate chat request, with trust_remote_code=False.
Preconditions: any non-harmony model (model_type != gpt_oss); renderer_num_workers at its default of 1; for the default-config authenticated API attacker vector, --chat-template must be unset (the default because most HF models ship their own).
Blast radius: API-server process, per replica. Engine/worker processes unaffected by the CPU variant (event loop survives); affected by the memory variant (process OOM).
Persistence: until process restart. Survives client disconnect.
This advisory was fixed in MR50300.
{
"affected": [
{
"package": {
"ecosystem": "PyPI",
"name": "vllm"
},
"ranges": [
{
"events": [
{
"introduced": "0.19.0"
},
{
"fixed": "0.32.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-115911"
],
"database_specific": {
"cwe_ids": [
"CWE-400",
"CWE-770",
"CWE-834"
],
"github_reviewed": true,
"github_reviewed_at": "2026-10-28T21:14:52Z",
"nvd_published_at": "2026-10-26T17:17:18Z",
"severity": "MODERATE"
},
"details": "### Summary\nvLLM renders chat templates by passing the resolved Jinja2 source to `tokenizer.apply_chat_template()`, which compiles and `.render()`s it in an `ImmutableSandboxedEnvironment`. The sandbox prevents code-execution escapes but imposes **no wall-clock timeout, no output-size cap, and no nested-loop or binop limit**. The render runs in a `ThreadPoolExecutor` whose `max_workers` defaults to **1**, and CPython worker threads cannot be interrupted. An authenticated API client (or a malicious model repo loaded with `trust_remote_code=False`) can therefore submit a single short template and permanently wedge **all** chat-style endpoints (`/v1/chat/completions`, `/v1/responses`, `/tokenize`, `/pooling`-chat, `/invocations`-chat) for every tenant until the process is restarted. A string-multiplication variant instead OOM-kills the API-server process. Because of a separately-reported (GHSA-jmjw-q4x3-hmc2) gate omission on `/v1/responses`, the remote authenticated attacker vector works in **default `vllm serve` configuration** with no operator opt-in.\n\n### Impact\nA single authenticated request of \u003c100 bytes permanently denies all chat-style inference (`/v1/chat/completions`, `/v1/responses`, `/tokenize`, `/pooling`-chat, `/invocations`-chat) to **every tenant** of the API-server replica until the operator restarts the process. The string-multiplication variant OOM-kills the process. In a multi-tenant deployment this is a cross-tenant denial-of-service from any one tenant. From the malicious-model angle, a model repo whose `tokenizer_config.json` carries the hostile template wedges every replica that loads it on the first legitimate chat request, with `trust_remote_code=False`.\n\n**Preconditions:** any non-harmony model (`model_type != gpt_oss`); `renderer_num_workers` at its default of 1; for the default-config authenticated API attacker vector, `--chat-template` must be unset (the default because most HF models ship their own).\n**Blast radius:** API-server process, per replica. Engine/worker processes unaffected by the CPU variant (event loop survives); affected by the memory variant (process OOM).\n**Persistence:** until process restart. Survives client disconnect.\n\nThis advisory was fixed in [MR50300](https://github.com/vllm-project/vllm/pull/50300).",
"id": "GHSA-542m-9234-8gvf",
"modified": "2026-10-28T21:14:52Z",
"published": "2026-10-28T21:14:52Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-542m-9234-8gvf"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-115911"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/pull/50300"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/commit/cff08b461e086523cee8bba5bd3c66d302625134"
},
{
"type": "PACKAGE",
"url": "https://github.com/vllm-project/vllm"
},
{
"type": "WEB",
"url": "https://github.com/vllm-project/vllm/releases/tag/v0.32.0"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"type": "CVSS_V3"
}
],
"summary": "vLLM: DoS via unbounded chat_template Jinja2 rendering exhausts the renderer thread pool"
}
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
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.