🛡 Cisco Introduces Antares for Vulnerability Localization

The company has released the Antares family of compact, open-weight models that specialize in locating vulnerabilities in code. The release includes the Antares-350M and Antares-1B models, which mimic the actions of a security researcher.

🌍 The use of compact SLMs allows for deep security auditing directly within on-premises environments without sending proprietary code to cloud-based LLMs, solving privacy and cost issues.

👤 Professional code auditing becomes available locally on standard hardware, simplifying AppSec tasks even for small teams and individual developers.

Source 1: https://blogs.cisco.com/ai/introducing-antares-the-most-efficient-open-weight-ai-models-for-vulnerability-localization