Inception

Inception Labs is the AI platform behind Mercury, a family of diffusion-based large language models (LLMs). Unlike traditional auto-regressive models that generate tokens sequentially, Mercury uses a diffusion process that produces many tokens in parallel, resulting in significantly faster inference speeds and lower operational costs—reportedly several times faster and less than half the cost of conventional LLMs. The platform is designed for chat, coding, and enterprise inference, and its diffusion framework offers fine-grained control over outputs, enabling adherence to specific schemas and semantic constraints. Based on the company's descriptions, the approach aims to unify language with other data modalities such as audio, images, and video, though current publicly available interaction appears primarily text-based. The team behind Inception includes researchers and engineers from leading institutions and companies, and the models are already being deployed at Fortune 500 enterprises.