BREAKING
Mercury 2 Goes Live on Design Arena
World's First Reasoning dLLM
0tok/s
Speed (official)
0$/1M
Input price
0$/1M
Output price
0K
Context length
Partner Gains in Production
Latency cut82
Cost cut90
Strengths vs Drawbacks
Strengths
Near-1,000 tokens/sec
OpenAI API compatible
Native tool use and JSON
Drawbacks
Medium to poor writing
Fabricated references
Weak emotional nuance
Can Diffusion LLMs Scale Up
AI NEWS BLITZ
Inception Labs brings its diffusion model Mercury 2 to the Design Arena platform.