River AI raises $1.1bn for custom model development platform

River AI, a full-stack AI company based in Palo Alto, California, has closed a $1.1bn funding round led by General Catalyst and AMP PBC, with strategic investment from Nvidia and AMD Ventures. Y Combinator and Temasek also participated. The company was founded by Igor Babuschkin, who previously served as co-founder of xAI and held roles in generative modeling and reinforcement learning at Google DeepMind and OpenAI, where he oversaw large-scale training efforts. The founding team includes people who previously worked at xAI and Tesla.

River AI's core offering is a platform and API that enables developers and enterprises to train, tune, and serve their own custom AI models. The API provides LoRA fine-tuning and reinforcement learning capabilities for frontier open weight models. Billing is metered on tokens used for training and inference—a model that the company says removes the cost of idle GPU capacity. Trained models can be deployed to production instantly. The platform handles fast weight transfers, sampling-training consistency, and elastic compute, addressing technical barriers that have historically required dedicated infrastructure teams, specialized hardware, and months of work.

According to River AI, an enterprise can complete a complex reinforcement learning training run in 15 to 20 minutes without needing an infrastructure team. The company claims cost savings of two to four times relative to closed-source alternatives. Babuschkin stated: "The way AI is built today is not how it will be built in the future. AI should be open, freely available, and affordable. It should feel like it is working for the person using it, not the lab that trained it. We started River to allow people and companies to own their intelligence." The company's stated aim is to place ownership of AI with the people and organizations using it.

River AI plans an integrated stack covering new hardware, training infrastructure, and products centered on personalization and continual learning. Its longer-term goal is to extend the same control to individual users. General Catalyst managing director Marc Bhargava said: "There is a gap between what AI can do and what most companies actually experience. Until now, companies have lacked a cost-efficient way to train, tune, and own custom AI models. River closes this gap, helping any company build models on their own data, tailored to how they actually work."