OpenAI Begins Testing Ultrafast Tier for GPT-5.6 Sol

As AI companies continue to improve model performance, OpenAI is putting more emphasis on speed. 

The company is testing Ultrafast, a new way to run GPT-5.6 Sol at much faster processing than OpenAI’s Standard option. Access is currently restricted through the OpenAI API as the company tests the added speed across more demanding applications.

For businesses, access to the added speed could make GPT-5.6 Sol more valuable as expectations for immediate, seamless digital experiences become the norm.

What OpenAI is Testing 

OpenAI announced that Ultrafast can run GPT-5.6 Sol up to 14 times faster than its Standard processing option.  At peak performance, it can reach 750 output tokens per second. This added speed is powered by OpenAI’s partnership with Cerebras, combining advanced computing infrastructure with Sol’s reasoning capabilities.

But OpenAI is testing more than how quickly Sol can generate answers. The larger question looming is whether frontier models can achieve faster response times without lowering the bar on their ability to handle complex tasks.

OpenAI developers are using Ultrafast in situations where timing can directly affect outcomes, including:

• Reviewing logs during system disruptions
• Researching information across multiple sources
• Handling multi-step customer requests
• Analyzing financial activity

A small group of companies outside OpenAI is also using Ultrafast. Their testing is being used to provide useful clues about where added speed may have the most value for complex business workflows.

Why Speed Is Shaping the AI Race

The AI race is no longer only focused on building smarter models. Speed is now a major part of the competition.

The stakes are rising as businesses rely on AI to handle more complicated tasks and expect reliable results quickly. Each additional step means more time before the user can reach a useful outcome.

Keeping response times low becomes more important as companies begin deploying AI agents for more complex business functions. 

These systems may need to pull information from multiple sources, use third-party tools to evaluate what they find, and decide what to do next. That puts pressure on companies to design systems that move faster without compromising the intelligence and performance businesses expect when handing tasks over to AI systems.

With Ultrafast, response time is another benchmark to watch. AI platforms will continue to work to build faster models without lowering the bar for what those models can deliver.

Get Ahead as AI Continues to Accelerate 

As AI takes on more of the work behind customer interactions in less time, businesses should see more opportunities to answer questions, solve problems, and guide decisions in real time. Speed can create an opportunity. 

But faster responses are only worthwhile if they are relevant and reflect accurate information about your brand. 

This is where AI Search Optimization (AEO) becomes a key part of the equation. As AI-generated responses become more immediate, brands need to pay close attention to how accurately they are presented across AI-driven discovery.

Ready to learn how your brand is showing up across AI search? Start with Avenue Z’s AEO Readiness Assessment to see where you can gain ground.