# OpenAI Achieves Major Breakthrough in AI Image Generation Speed

OpenAI announced a significant advancement in continuous-time consistency models, a technology that could dramatically accelerate AI image generation while maintaining quality.

The company revealed they have simplified, stabilized, and scaled these models to match the sample quality of leading diffusion models—but with a crucial difference. The new approach requires only two sampling steps, compared to the dozens or even hundreds typically needed by traditional diffusion models like those powering DALL-E and Stable Diffusion.

This matters because generating images with diffusion models is computationally expensive and time-consuming. Each sampling step requires significant processing power, making real-time generation challenging and costly. By reducing the process to just two steps while preserving quality, OpenAI has potentially solved one of the biggest bottlenecks in AI image generation.

The breakthrough could enable faster, more efficient AI image creation across applications—from creative