Inception

Inception is an AI research and product company pioneering diffusion-based large language models (dLLMs) as a faster alternative to traditional autoregressive LLMs. While conventional LLMs generate one token at a time sequentially, Inception's models use a coarse-to-fine parallel refinement process — the same architectural approach that powers image generation — to produce text at over 1,000 tokens per second, 5x faster than speed-optimized models at roughly half the cost. The company's flagship product, Mercury, is the world's first commercially available family of dLLMs, with Mercury Coder targeting software development workflows and Mercury 2 adding reasoning capabilities. Founded by Stanford professor Stefano Ermon (who co-invented diffusion models), UCLA professor Aditya Grover, and Cornell professor Volodymyr Kuleshov, the company is headquartered in Palo Alto and backed by Mayfield, M12, and notable angels including Eric Schmidt, Patrick and John Collison, Andrej Karpathy, and Andrew Ng.