EdgeRunner AI and the US Army’s Artificial Intelligence Integration Center (AI2C) have developed a military-focused large language model that reportedly reduces error rates by up to 37 percent across critical Army tasks.
Named EdgeRunner-Camo, the open-weight model is optimized for local deployment and can operate in air-gapped environments without depending on cloud-based AI infrastructure.
The development team analyzed 1.4 million CamoGPT conversations comprising 17.7 million messages. Non-military queries, personally identifiable information, and requests requiring external tools or retrieval were filtered out during the process.
