# CVE-2026-115914: vLLM denial of service in DeepSeek-OCR image processing

Source: https://junglewise.ai/threats/cve-2026-115914-vllm-is-an-inference-and-serving-engine-for-large-language  
Publisher: Junglewise Threat Intelligence, https://junglewise.ai/threats  
Published: 2026-10-26. Updated: 2026-10-26  
License: CC BY 4.0 with a link back

CVE-2026-115914 · Severity: medium · CVSS 6.5 · Published 2026-10-26

vLLM is an inference engine for large language models that supports vision capabilities. An attacker with network access to the OpenAI-compatible API can submit a specially crafted image whose pixel values sum to zero, causing the engine to crash and deny service to all other users. This requires only the ability to reach the API endpoint and send a valid image file.

Technologies: [vllm (PyPI)](https://junglewise.ai/threats/technologies/pypi-vllm), [Vllm](https://junglewise.ai/threats/technologies/vllm). Vendors: [PyPI](https://junglewise.ai/threats/vendors/pypi), [Vllm](https://junglewise.ai/threats/vendors/vllm).

## Quick answers

**Is CVE-2026-115914 exploited in the wild?** No. Junglewise Threat Intelligence has no report of CVE-2026-115914 being exploited in the wild.

**How severe is CVE-2026-115914?** Medium severity, with a CVSS score of 6.5, published 26 October 2026.

**Which products and versions are affected?** vllm (PyPI) >= 0.11.1, < 0.32.0, ECOSYSTEM:introduced=0.11.1,fixed=0.32.0; Vllm 0.11.1 to 0.31.x.

**Is there a fix for CVE-2026-115914?** Yes. vllm (PyPI): upgrade to 0.32.0 or later. Junglewise finds every place you run it, drafts the upgrade and applies it once you approve.

**How is CVE-2026-115914 attacked?** It is attacked over the network, in a software library.

**Am I affected by CVE-2026-115914?** Check whether you run vllm (PyPI) or Vllm at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

## How to fix

- vllm (PyPI): upgrade to 0.32.0 or later

  ```sh
  pip install "vllm>=0.32.0"
  ```


## Technical details

The vulnerability exists in the DeepSeek-OCR, DeepSeek-OCR-2, and Unlimited-OCR image path processing. When an image's normalized pixel tensor sums to exactly zero, _parse_and_validate_image_input returns None instead of an embedding, which propagates through embed_multimodal and triggers a fatal assertion in sanity_check_mm_encoder_outputs, crashing the entire inference engine. The attack requires network access to the API but no authentication or user interaction.

## Affected products

- vLLM vLLM 0.11.1 to 0.31.x

## Identifiers

- CVE-2026-115914
- GHSA-gq5q-89c7-vhvx

## Timeline

- 2026-10-26: disclosed
- 2026-10-02: patched: Fix merged in commit c15672c

## References

- [https://github.com/vllm-project/vllm/commit/c15672cfb49152ebf65cfd188e71b84593ec5686](https://github.com/vllm-project/vllm/commit/c15672cfb49152ebf65cfd188e71b84593ec5686)
- [https://github.com/vllm-project/vllm/pull/59417](https://github.com/vllm-project/vllm/pull/59417)

## Related threats

- [vLLM chat template resource exhaustion DoS](https://junglewise.ai/threats/cve-2026-115918-vllm-is-an-inference-and-serving-engine-for-large-language)
- [vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP server](https://junglewise.ai/threats/cve-2026-115917-vllm-is-an-inference-and-serving-engine-for-large-language)
- [vLLM off-by-one error in tokenizer max_token_id validation](https://junglewise.ai/threats/cve-2026-115916-vllm-is-an-inference-and-serving-engine-for-large-language)
- [vLLM image format validation bypass in media processing](https://junglewise.ai/threats/cve-2026-115915-vllm-is-an-inference-and-serving-engine-for-large-language)
- [vLLM Phi-4 multimodal audio encoder GPU memory exhaustion](https://junglewise.ai/threats/cve-2026-115913-vllm-is-an-inference-and-serving-engine-for-large-language)