What Inception does
Inception is an AI research and product company building diffusion-based large language models (dLLMs) for production applications where latency, throughput, and multi-step agent loops matter. On its website, the company contrasts diffusion-based generation (parallel coarse-to-fine refinement across a small number of steps) with autoregressive “token-by-token, left-to-right” decoding, arguing diffusion can enable a different speed–cost curve that better fits always-on, high-volume systems. Inception’s flagship model family is Mercury, which it positions as a commercially available diffusion LLM designed to integrate into existing LLM workflows and serve as a drop-in replacement in latency-sensitive deployments via an API and enterprise/on-prem deployment options.