NvidiaCVE-2026-47475
Nvidia TensorRT-LLM: denial of service
Medium6.2CVE-2026-47475 · Published Jul 14, 2026 · updated Jul 15, 2026
NVIDIA TensorRT-LLM contains a vulnerability in the OpenAI-compatible inference API where an attacker could trigger a reachable assertion in the sampler thread. A successful exploit of this vulnerability might lead to denial of service.
Affected versions
| Package | Affected | Fixed in |
|---|---|---|
| TensorRT-LLM Product | >= 0.0, <= v1.3.0 rc15 | No fix yet |
Details and references
- CVSS 3.1
- CVSS:3.1/AV:L/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
- Severity from
- the vendor (its own CVE record or advisory)
- Weakness
- CWE-617
More Nvidia advisories
All Nvidia| Date | Advisory | Severity | Fixed in |
|---|---|---|---|
| Jul 14 | Nvidia TensorRT-LLM: denial of service | Medium6.2 | No fix yet |
| Jul 14 | Nvidia TensorRT-LLM: unsafe deserialization | High7.5 | No fix yet |
| Jul 14 | Nvidia TensorRT-LLM: unsafe deserialization | High7.8 | No fix yet |
| Jul 14 | Nvidia TensorRT-LLM: tampering | High7.4 | No fix yet |
| Jul 14 | Nvidia TensorRT-LLM: information disclosure | High7.3 | No fix yet |
| Jul 14 | Nvidia TensorRT-LLM: unsafe deserialization | High8.4 | No fix yet |