OpenAI Releases Ultrafast Mode for GPT-5.6 Sol Model
AI company OpenAI has launched a preview of its new Ultrafast mode for the GPT-5.6 Sol model. This new mode operates up to 14 times faster than previous versions, generating up to 750 tokens per second.
AI research company OpenAI has announced the preview release of an 'Ultrafast' mode for its GPT-5.6 Sol model. The new mode is reported to be up to 14 times faster than standard processing, capable of generating a maximum of 750 tokens per second.
The Ultrafast mode has been developed with support from Cerebras, aiming to significantly improve response times within user workflows. Historically, achieving higher speeds in AI models often required a trade-off in capabilities or the use of smaller, more specialized models.
OpenAI states that this new mode allows for the simultaneous achievement of both speed and capacity. The company has identified several early use cases where this enhanced speed is beneficial, including accident response and reliability analysis, financial research and security analysis, customer service and voice interactions, e-commerce product inquiries and inventory checks, and real-time research and experimentation.
The Ultrafast mode is currently in a limited preview phase, accessible to a select group of clients. OpenAI is using this early testing to assess where the speed increase provides the most value and to observe how well the model can keep pace with user operations.