- GPT-5.6 Sol’s Ultrafast mode can process data up to 14 times faster than the standard version.
- The new mode generates up to 750 output tokens per second, enhancing productivity for developers.
- OpenAI is testing Ultrafast across various applications, including coding and financial research, to maximize speed benefits.
Read next
OpenAI explores sparse models to enhance AI debugging capabilities
November 17, 2025
OpenAI researchers are testing sparse models as a potential solution to improve the transparency and…
Apple Advances in AI Image Generation, Competing with DALL-E and Midjourney
June 10, 2025
Apple introduces STARFlow, an innovative AI image generation system that rivals diffusion models like DALL-E and…
Academic Publishers Turn to AI Partnerships Amid Research Funding Cuts
June 10, 2025
Academic publishers are inking licensing deals with AI firms to generate new revenue streams amid reduced…
Reddit sues Anthropic for not paying for training data
June 5, 2025
Reddit has filed a lawsuit against Anthropic for reportedly using Reddit's data to train AI models without a…