CVE-2026-116172: vLLM 0.22.0 before 0.24.0 DoS via Rust gRPC CandidateTokens.top_n

Q: Which deployments are exposed to this denial-of-service condition?

vLLM versions from 0.22.0 up to, but not including, 0.24.0 are affected when they expose the Rust gRPC converter and accept Generate or GenerateStream requests.

Q: What does an attacker need to do to trigger the failure?

An attacker can remotely submit a Generate or GenerateStream request containing an oversized or integer-wrapping CandidateTokens.top_n value. No privileges or user interaction are required, but exploitation has high attack complexity.

Q: What is the operational impact if exploitation succeeds?

The malformed top_n value can bypass max_logprobs limits and cause torch.topk failures that terminate EngineCore, resulting in denial of service. The provided information indicates no confidentiality or integrity impact.

Q: What should teams do if they cannot immediately upgrade?

The available information does not specify a workaround. Limit access to the affected gRPC request interface and prevent untrusted clients from sending Generate or GenerateStream requests until upgrading to version 0.24.0 or later.

Published Oct 25, 2026

Updated

vLLM from 0.22.0 before 0.24.0 contains an improper input validation vulnerability in the Rust gRPC converter that allows remote attackers to bypass maxlogprobs limits via CandidateTokens.topn. Attackers can send Generate or GenerateStream requests with oversized or integer-wrapping topn values to trigger torch.topk failures that terminate EngineCore.

Affected Software

1 affected component

vllm vllm>=0.22.0<0.24.0

Event History

Oct 25, 2026

CVE Published

via MITRE·01:20 PM

Data Sourced

via MITRE·01:20 PM

DescriptionSeverityWeakness

Frequently Asked Questions

Which deployments are exposed to this denial-of-service condition?

vLLM versions from 0.22.0 up to, but not including, 0.24.0 are affected when they expose the Rust gRPC converter and accept Generate or GenerateStream requests.

What does an attacker need to do to trigger the failure?

An attacker can remotely submit a Generate or GenerateStream request containing an oversized or integer-wrapping CandidateTokens.top_n value. No privileges or user interaction are required, but exploitation has high attack complexity.

What is the operational impact if exploitation succeeds?

The malformed top_n value can bypass max_logprobs limits and cause torch.topk failures that terminate EngineCore, resulting in denial of service. The provided information indicates no confidentiality or integrity impact.

What should teams do if they cannot immediately upgrade?

The available information does not specify a workaround. Limit access to the affected gRPC request interface and prevent untrusted clients from sending Generate or GenerateStream requests until upgrading to version 0.24.0 or later.

CVE-2026-116172: vLLM 0.22.0 before 0.24.0 DoS via Rust gRPC CandidateTokens.top_n

Affected Software

Event History

Frequently Asked Questions

Which deployments are exposed to this denial-of-service condition?

What does an attacker need to do to trigger the failure?

What is the operational impact if exploitation succeeds?

What should teams do if they cannot immediately upgrade?

Company

Resources

Contact

Frequently Asked Questions

Which deployments are exposed to this denial-of-service condition?

What does an attacker need to do to trigger the failure?

What is the operational impact if exploitation succeeds?

What should teams do if they cannot immediately upgrade?