medium6.5
CVE-2026-115914
vLLM is an inference and serving engine for large language models.
Impact
vLLM is an inference and serving engine for large language models. From 0.11.1 until 0.32.0, the DeepSeek-OCR, DeepSeek-OCR-2, and Unlimited-OCR image path allows a remote client that can reach the OpenAI-compatible API to submit a legal image whose normalized pixels sum to zero when that image is alone in its per-step encoder batch. _parse_and_validate_image_input returns no embedding, embed_multimodal propagates that value, and sanity_check_mm_encoder_outputs triggers a fatal assertion that shuts down the engine for all users. This issue is fixed in version 0.32.0.
Affected packages
| Ecosystem | Package | Affected | Fixed in |
|---|---|---|---|
| pip | vllm | >= 0.11.1, < 0.32.0 | 0.32.0 |
Affected products
| Vendor | Product | Affected versions |
|---|---|---|
| Vllm | Vllm | from 0.11.1 before 0.32.0 |
Weakness classification
References
- NVD record
- CVE.org record
- GitHub advisory
- https://github.com/advisories/GHSA-gq5q-89c7-vhvx
- https://github.com/vllm-project/vllm/security/advisories/GHSA-gq5q-89c7-vhvx
- https://nvd.nist.gov/vuln/detail/CVE-2026-115914
- https://github.com/vllm-project/vllm/pull/59417
- https://github.com/vllm-project/vllm/commit/c15672cfb49152ebf65cfd188e71b84593ec5686
- https://github.com/vllm-project/vllm/releases/tag/v0.32.0