Frontier model
One of the most capable models at a given time, usually from a handful of well-funded labs.
Frontier model is an informal label for the most capable systems of the moment, such as the latest from OpenAI, Anthropic, Google DeepMind, xAI and Meta, with DeepSeek and other Chinese labs close behind on many benchmarks. The term became common in policy after the 2023 AI Safety Summit at Bletchley Park, and several labs have published frontier safety frameworks describing tests they run before release. Because the frontier moves every few months, a frontier model of one year is often matched by an open model the next.