River AI, founded by xAI co-founder Igor Babuschkin just two months ago, has raised $1.

1 billion in a seed/Series A round led by General Catalyst and AMP PBC, with participation from Nvidia, AMD Ventures, Y Combinator, and Temasek.

Babuschkin, previously at DeepMind and OpenAI, aims to reinvent AI from scratch, focusing on how models are trained to turn agents into personally trainable assistants rather than human worker replacements.

He writes that the stack must be rebuilt end to end: training, models, product layer, and hardware so that personal AI lives close to the user.

River already offers an API billed per 1 million tokens, using open models with reinforcement learning (RL) and low-rank adaptation (LoRA) fine-tuning.

The product is positioned as an antidote to prompt engineering, letting users train open models into ones they own and serve as an endpoint.

River claims any enterprise can complete a complex RL run in 15–20 minutes with no infrastructure team, at two to four times cost savings versus closed-source alternatives.

The bigger vision is everyone having their own agents, trained by themselves. This mirrors the rise of personal locally-running agents like OpenClaw and Nvidia’s AI-capable PC hardware partnerships.

How River’s tech will differ remains to be seen, but it starts with a war chest of cash.