In February 2026, Inception Labs released Mercury 2, billed as the world's first reasoning dLLM (Diffusion Large Language Model). Using a parallel-refinement generation process, it tops 1,000 tokens per second and is now available across several platforms, including Design Arena.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.