CVE-2026-94626 (CVSS 7.5)

Description

vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process.

Is your site exposed to CVE-2026-94626?

Run a free security scan — no signup, results in seconds.

CVSS v3.1 Score

7.5

HIGH

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

EPSS — Exploit Prediction

0.0063

Probability of exploitation

0.49%

Percentile rank

EPSS estimates the probability that this vulnerability will be exploited in the wild within the next 30 days. A higher score means more likely to be exploited.

Weakness Type (CWE)

CWE-789 CWE-789

References

Other References

https://github.com/vllm-project/vllm https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/utils.py#L569-L573 https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/metadata.py#L277 https://github.com/vllm-project/vllm/pull/51137 https://www.vulncheck.com/advisories/vllm-through-0.29.0-memory-exhaustion-via-unvalidated-nixl-tp-size

Frequently Asked Questions

What is CVE-2026-94626? +

vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process. It has a CVSS v3.1 base score of 7.5 (HIGH).

How severe is CVE-2026-94626? +

CVE-2026-94626 has a CVSS v3.1 score of 7.5 out of 10, rated HIGH. This is a high-severity vulnerability that should be prioritized for patching. The EPSS score is 0.0063, placing it in the 0th percentile for exploitation probability.

How do I check if I'm vulnerable to CVE-2026-94626? +

You can use Secably's free Website Scanner to check your website for known vulnerabilities. For infrastructure scanning, use the Port Scanner to identify exposed services that may be affected. Check the vendor advisories linked above for specific patch and version information.

Related Vulnerabilities

CVE-2026-54700

ConnectBot SSH Client Library is written in Kotlin with Kaitai Struct and KStateMachine. Prior to 0.3.1, the SSH protocol parser …

CVE-2026-63566

Memory allocation with excessive size value in the DTLS handshake reassembly (DtlsReliableHandshake, DtlsReassembler) in Legion of the Bouncy Castle Inc. …

CVE-2026-12185

In Bouncy Castle for Java before 1.85, BKS/UBER keystore allocates from untrusted lengths before integrity check. This issue also affects …

CVE-2026-58060

In Bouncy Castle for Java before 1.85, HSS public-key level count unbounded, enabling huge allocation on verify. This issue also …

CVE-2026-12852

In Bouncy Castle for Java before 1.85, MLS wire decoder allocates attacker-declared opaque length before bounds check.

CVE-2026-14682

In Bouncy Castle for Java before 1.85, Possible OOM from unbounded up-front allocation on a definite-length read. This issue also …