EdgeRunner AI and the US Army’s Artificial Intelligence Integration Center (AI2C) have created a military-specific large language model that reportedly cuts error rates by up to 37 percent across key army tasks.
Called EdgeRunner-Camo, the open-weight model is fine-tuned to run locally, including in air-gapped environments without relying on cloud-hosted AI.
EdgeRunner and AI2C analyzed 1.4 million CamoGPT conversations containing 17.7 million messages, filtering out non-military queries, personally identifiable information, and questions requiring external tools or retrieval.
Read the full story on Military AI: US Army, EdgeRunner Develop Military AI Model That Cuts Errors by Up to 37%