What does this diffusion-based AI model do differently? Yahoo Finance Video Mon, September 14, 2026 at 11:18 p.m. GMT+2 Inception founder and CEO Stefano Ermon — whose company just launched its Mercury 2.5 AI model — explains how the diffusion-based AI model differs from other sorts of large language models (LLMs). Terms and Privacy Policy EU DSA Contact Privacy & Cookie Settings
Inception ships Mercury 2.5 at 1,107 tokens per second, by its count