CVE-2026-94626 vLLM through 0.29.0 Memory Exhaustion via Unvalidated NIXL tp_size
Attackers can cause denial of service in vLLM by sending unvalidated tp_size parameters to exhaust memory in disaggregated deployments.
- Exploited in the wild
- Not known
- Login needed
- Not needed
- Attackable
- Over the internet
- Victim action needed
- Not needed
- Fix
- Available
What is CVE-2026-94626?
CVE-2026-94626 is a high-severity denial of service vulnerability in Vllm Vllm, published on 21 Sep 2026. An attacker can exploit it over the internet without logging in, with no action from the victim.
Official 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.
Weakness: CWE-789
Which versions are affected by CVE-2026-94626?
1 open source package ship an affected version, including vllm. It affects 1 product.
| Package | Affected | Fixed in |
|---|---|---|
vllm PyPI | All versions | 0.30.0 |
Products include Vllm Vllm.
How do I fix CVE-2026-94626?
Upgrade vllm to 0.30.0 or later.
Upgrade, then put limits on request size, rate and nesting depth so one client cannot exhaust the service. More about Denial of service.
Is CVE-2026-94626 being exploited?
There is no known exploitation of CVE-2026-94626 in the wild and no public exploit code.
Other vulnerabilities in Vllm Vllm
- CVE-2025-32444 vLLM Vulnerable to Remote Code Execution via Mooncake Integration
- CVE-2026-22778 vLLM leaks a heap address when PIL throws an error
- CVE-2025-47277 vLLM Allows Remote Code Execution via PyNcclPipe Communication Service
- CVE-2024-11041 Remote Code Execution in vllm-project/vllm
- CVE-2026-48746 vLLM: OpenAI auth bypass
Sources
Also known as GHSA-hq63-vm56-2666, MINI-7x93-9hg9-r4rv, MINI-963p-63fx-xjgc, MINI-f78c-cgfc-2wqh, PYSEC-2026-4008.