Powered by @Cerebras, Ultrafast generates up to 750 tokens per second, bringing

Frontier-quality output at ~750 tokens/second changes what fits inside an interactive loop — real-time voice agents and tight coding iterations that currently stall on generation speed.

• token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.

• videoProduction Monitoring with Codex: Grafana, Kubernetes, & Security
• videoMore visual answers from GPT-6 with Intelligent UI
• videoModel Behavior: The Science of AI Style
• videoMeasuring Agents With Interactive Evaluations

• articlePreviewing Ultrafast mode: GPT-5.6 Sol at up to 14X the speedopenai.com
• postPreviewing Ultrafast mode: GPT-5.6 Sol at up to 14x the speed. Launching first iOpenAI
• postOpenAI Ultrafast makes Astra up to 8x faster in Codex and the APIOpenAI Developers

Checking sign-in…

Loading comments…