CVE-2025-15514 Ollama CVE debrief

Who should care

Organizations using Ollama versions between 0.11.5-rc0 and 0.13.5 should prioritize patching this vulnerability to prevent potential denial of service attacks. Additionally, security teams and administrators responsible for maintaining Ollama installations should be aware of this vulnerability and take necessary actions to mitigate the risk.

Technical summary

The vulnerability exists in the multi-modal model image processing functionality of Ollama, specifically in the handling of base64-encoded image data via the /api/chat endpoint. The application fails to validate the decoded data before passing it to the mtmd_helper_bitmap_init_from_buf function, which can return NULL for malformed input. The code does not check this return value before dereferencing the pointer, leading to a null pointer dereference and a segmentation fault. This results in a denial of service condition, making the model unavailable to all users until the service is restarted.

Defensive priority

High priority should be given to patching this vulnerability, as it can be exploited remotely and results in a denial of service condition. Administrators should update Ollama to a version that fixes this issue as soon as possible.

Recommended defensive actions

• Update Ollama to a version that fixes the null pointer dereference vulnerability.
• Implement input validation and sanitization for base64-encoded image data.
• Monitor Ollama installations for potential exploitation attempts.
• Review and update incident response plans to address potential denial of service attacks.
• Consider implementing compensating controls, such as web application firewalls, to detect and prevent malicious traffic.

Evidence notes

The CVE-2025-15514 vulnerability was published on January 12, 2026, and last modified on June 30, 2026. The vulnerability affects Ollama versions from 0.11.5-rc0 to 0.13.5. The CVSS score is 8.7, indicating high severity. The vulnerability allows remote attackers to exploit the application, causing a denial of service condition.

Sources and references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.