This is our second edition of our quarterly report on the capabilities of generative AI models.

The AI bubble has gone from a topic of national discourse, to distinctly affecting private markets. Q2 saw new generative AI startups raise billions of dollars in capital, existing labs come up iteratively better models, and the standard of frontier model usage moving to reasoning and research. The default model for all major model companies in the United States is now a reasoning model, showcasing the speed of development of contemporary AI applications.

Download PDF or read the full report here 👇