CVE-2026-116173: vLLM 0.22.0 before 0.24.0 Denial of Service via gRPC output_candidates.token_ids

Q: Which deployments are exposed to this issue?

vLLM versions from 0.22.0 up to, but not including, 0.24.0 are affected when the Rust gRPC frontend accepts Generate or GenerateStream requests. An attacker can reach the vulnerable path remotely without authentication.

Q: What request conditions are required to trigger the denial of service?

The attacker must send a Generate or GenerateStream request with output_logprobs enabled and more than 128 values in output_candidates.token_ids. This bypasses the logprob_token_ids limit and causes a fatal sampler exception that terminates EngineCore.

Q: What is the operational impact of successful exploitation?

Successful exploitation crashes EngineCore, causing denial of service. The provided information identifies no confidentiality or integrity impact.

Published Oct 25, 2026

Updated

vLLM from 0.22.0 before 0.24.0 contains a denial of service vulnerability that allows remote attackers to terminate EngineCore by bypassing logprobtokenids limits in the Rust gRPC frontend. Attackers can send Generate or GenerateStream requests with outputlogprobs enabled and more than 128 outputcandidates.tokenids, triggering a fatal sampler exception that crashes the engine.

Affected Software

1 affected component

vllm vllm>=0.22.0<0.24.0

Remediation

Recommended actions to resolve this vulnerability, in priority order.

Upgrade
Upgrade vLLM to a version that resolves this vulnerability.
Fixed in 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 issue?

vLLM versions from 0.22.0 up to, but not including, 0.24.0 are affected when the Rust gRPC frontend accepts Generate or GenerateStream requests. An attacker can reach the vulnerable path remotely without authentication.

What request conditions are required to trigger the denial of service?

The attacker must send a Generate or GenerateStream request with output_logprobs enabled and more than 128 values in output_candidates.token_ids. This bypasses the logprob_token_ids limit and causes a fatal sampler exception that terminates EngineCore.

What is the operational impact of successful exploitation?

Successful exploitation crashes EngineCore, causing denial of service. The provided information identifies no confidentiality or integrity impact.

CVE-2026-116173: vLLM 0.22.0 before 0.24.0 Denial of Service via gRPC output_candidates.token_ids

Affected Software

Remediation

Event History

Frequently Asked Questions

Which deployments are exposed to this issue?

What request conditions are required to trigger the denial of service?

What is the operational impact of successful exploitation?

Company

Resources

Contact

Frequently Asked Questions

Which deployments are exposed to this issue?

What request conditions are required to trigger the denial of service?

What is the operational impact of successful exploitation?