More than 40x the parameters of Mercury 1, still running at 1,000+ tokens/sec.
Congrats to the @_inception_ai team on Mercury 3!
Inception releases Mercury 3, a diffusion LLM running at 1,000+ tokens per second
AISummary
Modal congratulates Inception on Mercury 3, which it says has more than 40 times the parameters of Mercury 1 while running at 1,000+ tokens per second. Quoted post from Stefano Ermon says Mercury 3 is the company's most powerful model and the most intelligent diffusion LLM released so far. It says Mercury 3 outperforms GPT-6 Luna and Haiku 5.5 on quality while keeping that speed.
Post on XView on X
@modal
1/4 Today, we’re releasing Mercury 3. Mercury 3 is by far our most powerful model and the most intelligent diffusion LLM ever released. It outperforms GPT-6 Luna and Haiku 5.5 on quality while running at 1000+ tok/sec.
Source: Modal · x.comPublished · added here