OpenAI previews Ultrafast API tier for GPT-5.6 Sol
OpenAI is testing a Cerebras-powered API tier that pushes GPT-5.6 Sol to as much as 750 output tokens per second.
The Ultrafast preview is limited to selected API customers, with broader access planned as capacity grows. OpenAI frames it for latency-sensitive work in incident response, finance, security, support, voice, commerce, and research. The company says it is already using the tier internally for incident analysis and research workflows, while engineers still make deployment decisions. Pricing and general availability have not been announced. TestingCatalog's note
The Ultrafast preview is limited to selected API customers, with broader access planned as capacity grows. OpenAI frames it for latency-sensitive work in incident response, finance, security, support, voice, commerce, and research. The company says it is already using the tier internally for incident analysis and research workflows, while engineers still make deployment decisions. Pricing and general availability have not been announced. TestingCatalog's note
score 7