River AI secures $1.1 billion to rebuild the AI stack

River AI, a two-month-old startup founded by former xAI co-founder Igor Babuschkin, has secured $1.1 billion in a seed and Series A funding round. General Catalyst and AMP PBC—an investment firm founded in 2026 by former Andreessen Horowitz partner Anjney Midha—led the massive investment. Other participants in the round include Nvidia, AMD Ventures, Y Combinator, and Temasek. Babuschkin, who previously worked at DeepMind and OpenAI, launched the company in June with the goal of rebuilding the entire artificial intelligence stack.

The startup has already launched an API that charges developers per 1 million tokens, with pricing varying based on the specific open-weight model selected. This API serves as an alternative to traditional prompt engineering by allowing developers to apply reinforcement learning and low-rank adaptation fine-tuning directly to open models. According to the company, this approach lets developers transform open models into highly customized, proprietary endpoints.

For enterprise practitioners, River AI offers a neocloud environment designed to simplify post-training customization. The company claims that any enterprise can complete a complex reinforcement learning run in 15 to 20 minutes without needing a dedicated infrastructure team. This setup reportedly delivers two to four times the cost savings compared to closed-source alternatives, addressing the growing demand among businesses to control their own AI models.

Ultimately, River AI aims to shift the industry away from building human worker replacements and toward creating highly personalized, locally running digital assistants. This vision aligns with the broader rise of local agent frameworks like OpenClaw and hardware partnerships between Nvidia and PC manufacturers such as Dell, Microsoft, and HP. Backed by its new capital, River AI plans to redesign training, models, and the product layer to make personal AI a reality.