OpenAI introduces 'Ultrafast' mode for GPT-5.6 Sol

OpenAI has announced a new preview mode for its most powerful model, GPT-5.6 Sol, designed to dramatically accelerate response times for enterprise users. Dubbed 'Ultrafast,' this mode leverages Cerebras hardware to deliver up to 750 output tokens per second—a 14x speed increase over standard performance. The move is a direct response to growing demand for high-throughput AI in production environments, where latency and throughput can make or break real-time applications. For businesses running complex data pipelines, customer support automation, or large-scale content generation, the speed boost could translate into faster decision-making and lower operational costs. OpenAI is positioning this as a competitive edge, especially against rivals like Google and Anthropic, who are also pushing performance boundaries. The Ultrafast mode is currently in preview, meaning it may not be available to all users immediately, but the company has signaled that broader rollout is likely if early adoption goes well. Industry analysts see this as a strategic move to lock in enterprise contracts, where reliability and speed are often more important than raw capability. By partnering with Cerebras, OpenAI also diversifies its infrastructure dependencies, reducing reliance on Nvidia's dominant GPU ecosystem. While the pricing for Ultrafast mode has not been fully disclosed, the value proposition is clear: for workloads that demand near-instant responses, this could be a game-changer. As AI models become more integrated into daily business operations, the ability to process requests at unprecedented speeds will likely become a key differentiator in the market.