Vulnerability Description
vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the /v1/completions/derender and /v1/chat/completions/derender endpoints accept caller-supplied GenerateResponse objects whose generate_responses, choices, token_ids, prompt_logprobs, logprobs.content, top_logprobs, and routed_experts structures are processed by OnlineDerenderer and tokenizer.decode before max_model_len, max_tokens, max_num_seqs, or response-size limits are enforced, allowing an authenticated API client to consume excessive CPU and memory and produce oversized responses. This issue is fixed in version 0.26.0.
CVSS Score
MEDIUM
Related Weaknesses (CWE)
References
- https://github.com/vllm-project/vllm/commit/8e61b646e2d157f9b93451fa048f9c8530c8
- https://github.com/vllm-project/vllm/pull/47260
- https://github.com/vllm-project/vllm/releases/tag/v0.26.0
- https://github.com/vllm-project/vllm/security/advisories/GHSA-8737-qx52-hjff
FAQ
What is CVE-2026-71486?
CVE-2026-71486 is a vulnerability with a CVSS score of 4.3 (MEDIUM). vLLM is an inference and serving engine for large language models. Prior to 0.26.0, the /v1/completions/derender and /v1/chat/completions/derender endpoints accept caller-supplied GenerateResponse obj...
How severe is CVE-2026-71486?
CVE-2026-71486 has been rated MEDIUM with a CVSS base score of 4.3/10. Review the CVSS metrics above for detailed severity breakdown.
Is there a patch for CVE-2026-71486?
Check the references section above for vendor advisories and patch information. Review vendor security bulletins for remediation guidance.