Technology News and Headlines

Inception Launches Mercury 2, the First Diffusion-Based Language Reasoning Model

Inception Labs unveiled Mercury 2, the first diffusion‑based language reasoning model, claiming dramatic speed and cost advantages over leading models. The model generates 1,009 tokens per second with 1.7‑second end‑to‑end latency, beating Gemini 3 Flash and Claude Haiku on latency while delivering comparable benchmark scores. Pricing is $0.25 per million input tokens and $0.75 per million output tokens, roughly half to a quarter of competitors’ rates. Mercury 2 supports a 128K context window, tool usage, and JSON output, targeting latency‑sensitive enterprise use cases.