Arcee AI trained four models for $20 million. Now, it’s worth $1 billion.

In 2025, Mark McQuade founded Arcee AI, a startup focused on post-training AI models. However, Meta's retreat from open-weight models created an opportunity for Arcee. Open-weight models enable secure AI usage without sharing data with tech giants like OpenAI or Anthropic. To capitalize on this gap, McQuade committed $30 million of Arcee's $30 million capital to train four open-weight models, including a massive 400-billion-parameter model called Trinity Large.

Despite the relatively low cost of $20 million for training these models, Arcee has gained significant traction. The startup released a report indicating that it has raised a Series B round at a $1 billion pre-money valuation, led by Vista Equity Partners, Cambium Capital, and Emergence Capital, with participation from Microsoft's M12, AI10 Ventures, Hitachi, IAG, P7, and Wipro.

The funding will be used to develop new open-weight models, products, and grow partnerships with the U.S. Department of Energy and Vista's portfolio companies. McQuade's ambition is clear: to catch up to China in the AI race, recognizing that while the U.S. leads in closed-source AI, it has lagged behind in open-source AI development.