Inception Labs introduced Mercury 2, saying the model can generate about 1,000 tokens per second and scored 90% on AIME 2026.
According to NS3.AI, Google’s DiffusionGemma reaches similar generation speeds but scored 69.1% on the same benchmark.
AI TRENDS | Inception Labs Unveils Mercury 2 With 1,000 Tokens-Per-Second Output and 90% AIME 2026 Score
2026-06-21 16:04:31
Disclaimer:
1. The information provided does not constitute investment advice. Investors should make independent decisions and bear all risks themselves.
2. The copyright of this content belongs to the original author. The views expressed herein are solely those of the author and do not represent the stance or position of this website.
Previous article：
地缘政治 | 伊朗峰会回避与美方合影
