Junglewise Threat Intelligence

CVE-2026-55514: vLLM denial of service via reachable assertion in M-RoPE models

CVE-2026-55514 · Severity: high · CVSS 4 · Published 2026-07-06

Technologies: vLLM Project vllm, vllm (PyPI). Vendors: vLLM Project, PyPI.

Executive brief

vLLM is a high-throughput engine for serving large language models. A vulnerability exists where an authorized user can crash the entire server by sending a specific type of request to the completions endpoint when using models that support Multimodal Rotary Positional Embedding (M-RoPE). This results in a complete denial of service, requiring a manual restart of the application to restore operations.

Technical details

A denial of service vulnerability exists in vLLM's GPUModelRunner due to an incorrect assertion in the _init_mrope_positions method. When a request is sent to the /v1/completions endpoint with prompt_embeds provided but prompt=None, the EngineCore fails a type-narrowing assertion that expects prompt_token_ids to be present. This failure triggers an unhandled AssertionError, causing the EngineCore and the APIServer to crash. The issue affects configurations where --enable-prompt-embeds is active and an M-RoPE-supported model is in use. A fix is available in version 0.24.0 which replaces the fatal assertion with logic to handle embedding-only inputs using dummy token IDs.

Affected products

  • vllm-project vllm >= 0.12.0, < 0.24.0

Timeline

  • 2026-07-02: disclosed
  • 2026-07-20: advisory: Last updated date

References

Related threats