Skip to content
Sign in
4.8 Medium

CVE-2026-90713 A security flaw has been discovered in vllm-project vLLM up to 0.29.0

Description

A security flaw has been discovered in vllm-project vLLM up to 0.29.0. The affected element is the function TiktokenTokenizer::new of the file rust/src/text/src/backend/hf/mod.rs of the component tiktoken vocab File Handler. The manipulation results in denial of service. The attack is only possible with local access. The exploit has been released to the public and may be used for attacks. The pull request to fix this issue awaits acceptance.

Affected products and versions

Product Affected versions Fixed in
vllm 0.29.0 No fix stated
vllm 0.28 No fix stated
vllm 0.27 No fix stated
vllm 0.26 No fix stated
vllm 0.25 No fix stated
vllm 0.24 No fix stated
vllm 0.23 No fix stated
vllm 0.22 No fix stated
vllm 0.21 No fix stated
Show the 20 other branches
Product Affected versions Fixed in
vllm 0.20 No fix stated
vllm 0.19 No fix stated
vllm 0.18 No fix stated
vllm 0.17 No fix stated
vllm 0.16 No fix stated
vllm 0.15 No fix stated
vllm 0.14 No fix stated
vllm 0.13 No fix stated
vllm 0.12 No fix stated
vllm 0.11 No fix stated
vllm 0.10 No fix stated
vllm 0.9 No fix stated
vllm 0.8 No fix stated
vllm 0.7 No fix stated
vllm 0.6 No fix stated
vllm 0.5 No fix stated
vllm 0.4 No fix stated
vllm 0.3 No fix stated
vllm 0.2 No fix stated
vllm 0.1 No fix stated

Other CVEs of vllm

The 121 CVEs of vllm →
CVE-2026-115918 vLLM is an inference and serving engine for large language models 6.5 Medium CVE-2026-115917 vLLM is an inference and serving engine for large language models 3.4 Low CVE-2026-115916 vLLM is an inference and serving engine for large language models 6.5 Medium CVE-2026-115915 vLLM is an inference and serving engine for large language models 8.1 High
Track this CVE