FlagThis

Abstract

In September 2025, the Center for AI Standards and Innovation (CAISI) at the National Institute of Standards and Technology (NIST) conducted a technical evaluation of three DeepSeek models (R1, R1-0528, and V3.1) and four U.S. reference models (OpenAI's GPT-5, GPT-5-mini, gpt-oss, and Anthropic's Opus 4). The assessment utilized 19 benchmarks across multiple domains, including public and private datasets, to evaluate capabilities, security robustness, cost efficiency, and alignment with foreign political narratives.