Kimi K3: Open-Weight AI Excels in Code Security

Aikido Security has integrated Moonshot AI's new open-weight model, Kimi K3, into its benchmark of large language models. This benchmark tests how well models can find known vulnerabilities. Kimi K3 rediscovered 23 out of 26 vulnerabilities, matching top proprietary models but at a lower cost. It also outperformed another open-weight model in the comparison. This suggests Aikido Security is exploring cost-efficient AI models for its security analysis. The use of open-weight systems could improve profit margins and reduce reliance on expensive proprietary providers. This trend might also give Aikido Security more flexibility in its model choices, potentially strengthening its market position.