Threats Tagged 'cve-2026-100653'

vLLM versions 0.22.1 through 0.28.0 have a vulnerability where the operator-supplied model revision pin (--revision / --code-revision) is not correctly propagated to certain Hugging Face artifact loads for FunAudioChat and Tarsier2 architectures. This causes deployments pinned to a specific revision to still fetch some processor, tokenizer, and configuration artifacts from the repository's default revision, potentially altering behavior without operator intent. This issue affects supply-chain integrity and reproducibility but does not allow remote code execution or bypass trust_remote_code settings. The vulnerability is fixed in version 0.28.0.

vLLM is an inference and serving engine for large language models. In versions from 0.22.1 through 0.28.0, the operator-supplied model revision pin (--revision / --code-revision) is not propagated to several Hugging Face artifact loads for the FunAudioChat and Tarsier2 architectures: the WhisperFeatureExtractor and speech_tokenizer PreTrainedTokenizerFast loads in vllm/model_executor/models/funaudiochat.py and the Qwen2VLConfig.from_pretrained call used by Tarsier2ProcessingInfo in vllm/model_executor/models/qwen2_vl.py. As a result, deployments pinned to a reviewed revision still resolve these behavior-affecting processor, tokenizer, and config artifacts from the repository's default revision, so a later change to the upstream default branch can alter audio preprocessing, speech tokenizer behavior, or Tarsier2 configuration without any change to the operator's configured pin. This is a supply-chain integrity and reproducibility failure for pinned deployments; it is residual to the earlier fix tracked as GHSA-3ww4-5jv9-j5gm / CVE-2026-47155 and does not constitute remote code execution or a trust_remote_code=False bypass. The issue is fixed in version 0.28.0.