OpenAI GPT-5.6 Sol Ultrafast mode preview

OpenAI is previewing Ultrafast mode for its GPT-5.6 Sol model, delivered through a partnership with chip company Cerebras, claiming speeds up to 14x faster than standard processing and up to 750 output tokens per second. The preview is available via API to select customers, with access set to expand as capacity grows. Cerebras published benchmark results showing Ultrafast completed 2,500 Humanity's Last Exam questions in 11 hours 11 minutes, compared to 78 hours 27 minutes for Claude Fable 5, with comparable accuracy. On GDP-Val, a benchmark for economically valuable knowledge-work tasks, Cerebras reported a 5.6x end-to-end speedup with no quality degradation. OpenAI is targeting the mode at corporate use cases including incident response, customer service, financial market analysis, and e-commerce, and is working with initial customers to understand where the speed improvement matters most. Competitors including Anthropic have launched similar accelerated modes; OpenAI claims greater speed than Claude's fast mode.