OpenAI shipped GPT-5.6 alongside a new Ultrafast API service tier powered by Cerebras infrastructure, capable of delivering up to 750 output tokens per second—a 14× performance boost over standard offerings. The acceleration addresses a critical friction point for AI applications requiring real-time responsiveness, particularly in agentic workflows where latency compounds across multiple API calls. This infrastructure advancement directly enables the faster, more cost-efficient agent building that startups and enterprises increasingly demand, positioning OpenAI to compete more aggressively in the inference optimization market.
The model release includes enhanced Responses API capabilities designed specifically for agent developers, streamlining the creation of autonomous systems. OpenAI's research indicates enterprises are actively transitioning from ChatGPT-as-assistant toward agentic AI execution, with frontier firms pulling ahead in deployment maturity. The GPT-5.6 feature set—particularly improved model selection and smarter routing—directly addresses this adoption curve, enabling builders to optimize performance-to-cost tradeoffs at scale without sacrificing quality.
Concurrently, OpenAI appointed Dali Rajic as Chief Revenue Officer to lead its global revenue organization and help enterprises realize full value from AI deployments. The executive move signals OpenAI's commitment to professional services, customer success, and enterprise-grade support infrastructure—critical differentiators beyond raw model capability. Combined with the infrastructure improvements and agentic AI focus, the CRO appointment indicates OpenAI is systematically positioning itself as an enterprise-grade AI platform provider rather than primarily a consumer-facing AI company.