Mercury Articles - Developers Digest

All TopicsMercuryLLM AI Diffusion Inception Labs Diffusion Models AI Architecture

LATEST

Mercury 2: The LLM That Doesn't Generate Like an LLM

Inception Labs shipped the first reasoning model built on diffusion instead of autoregressive generation. Over 1,000 tokens per second, competitive benchmarks, and a fundamentally different approach to how AI generates text.

February 24, 2026•8 min read

Read Article

Diffusion Language Models: How Mercury Changed the LLM Speed Game

7 min read

Inception Labs launched Mercury, the first commercial-grade diffusion large language model. It generates over 1,000 tokens per second on standard Nvidia hardware by replacing autoregressive generation with a coarse-to-fine diffusion process.

Diffusion Models Mercury LLM AI Architecture

MERCURY

Mercury 2: The LLM That Doesn't Generate Like an LLM

Diffusion Language Models: How Mercury Changed the LLM Speed Game

Get Smarter About AI Dev

MERCURY

Mercury 2: The LLM That Doesn't Generate Like an LLM

Diffusion Language Models: How Mercury Changed the LLM Speed Game

Get Smarter About AI Dev