Latest OpenAI, Anthropic Models Signal Incremental Progress Over Breakthrough, Arena Chief Says

Peter Gostev, head of AI capabilities at Arena, the community-based model-ranking platform, assessed this week that the latest releases from OpenAI and Anthropic represent refinement rather than frontier breakthroughs. Speaking to the accelerating release cadence of frontier models—OpenAI deployed GPT-6 Sol and Luna on September 22, just 19 days after GPT-6 Astra, while Anthropic released Claude Opus 5.5 the same day—Gostev noted that the distinguishing feature of a genuine frontier advance is that "you're going to feel it." The current wave of model updates, he argued, is simply "the next iteration."