Announcement
Mercury 3 launches with claimed 1,000-plus tokens per second
The launch post claims the diffusion LLM outperforms GPT-6 Luna and Haiku 5.5 on quality.
TLDR
The post announcing Mercury 3 calls it “by far our most powerful model” and “the most intelligent diffusion LLM ever released.” It claims the model outperforms GPT-6 Luna and Haiku 5.5 on quality while running at 1,000-plus tokens per second. Artificial Analysis cautions that the announcement’s AA-Omniscience and GDPval-AA v2 scores do not follow its methodology and are not comparable to its scores.
Combined views
659.1K
44 Sources, first seen ago
