# OpenAI GPT-5 Ultrafast mode launch

> Live situation record from CLSTR: https://clstr.news/situations/openai-gpt-5-ultrafast-mode-launch
> Updated: 2026-08-16T04:13:00.000Z. Sources: 32. Developments: 2.

OpenAI has introduced ‘Ultrafast’, a new API service tier designed to run its GPT-5. 6 Sol model at significantly higher speeds. Powered by Cerebras’ Wafer-Scale Engine technology, the mode can generate up to 750 output tokens per second, which is up to 14 times faster than standard processing.

This advancement aims to eliminate the traditional trade-off between model intelligence and response speed by allowing access to the full reasoning capabilities of GPT-5. 6 Sol with ultra-low latency. The service is currently in a limited preview for a select group of API customers in sectors such as financial research, coding, customer support, and e-commerce.

In performance benchmarks, Cerebras reported that the Ultrafast mode completed ‘Humanity’s Last Exam’—a 2,500-question PhD-level test—in approximately 11 hours and 11 minutes, which was notably faster than comparable performance from Anthropic’s Claude models. OpenAI is also utilizing the tier internally for time-sensitive tasks like incident response and log analysis.

## Claims

- OpenAI has introduced a new API service tier called Ultrafast for the GPT-5. 6 Sol model. (corroborated by 3 sources)
- The Ultrafast mode can generate up to 750 output tokens per second. (corroborated by 3 sources)
- The Ultrafast service is powered by hardware from Cerebras. (corroborated by 3 sources)
- Ultrafast uses the same GPT-5. 6 Sol model weights and capabilities rather than a smaller specialized model. (corroborated by 2 sources)
- The service is currently in a limited preview available to a small group of customers via the OpenAI API. (corroborated by 2 sources)

## Timeline

### 2026-08-16: OpenAI launches Ultrafast tier for GPT-5. 6 Sol

OpenAI has launched Ultrafast, a new API tier for GPT-5. 6 Sol powered by Cerebras chips, capable of generating up to 750 tokens per second—up to 14 times faster than standard processing.

6 sources. https://clstr.news/cluster/openai-launches-ultrafast-tier-for-gpt-5-6-sol

### 2026-08-13: OpenAI launches Ultrafast mode for GPT-5. 6 Sol powered by Cerebras

OpenAI has launched Ultrafast mode for GPT-5. 6 Sol, powered by Cerebras technology. The new API tier delivers up to 750 tokens per second, making it 14x faster than standard processing for real-time tasks.

27 sources. https://clstr.news/cluster/cerebras-powers-openais-gpt-5-6-sol-in-new-ultrafast-mode

---
Cite as: OpenAI GPT-5 Ultrafast mode launch. CLSTR, https://clstr.news/situations/openai-gpt-5-ultrafast-mode-launch