OpenAI’s new Ultrafast mode runs GPT-5.6 Sol 14 times faster, on Cerebras chips

OpenAI’s new Ultrafast mode runs GPT-5.6 Sol 14 times faster, on Cerebras chips

OpenAI previewed Ultrafast, running GPT-5.6 Sol at 14x speed on Cerebras chips (750 tokens/sec), eliminating the intelligence-latency trade-off for agentic AI. Latency becomes the go-to-market moat—inference-optimized architectures now displace training capacity as the market differentiator, signaling diversification away from Nvidia-only dependencies.