River AI Raises $1.1B to Build Personally Trainable AI Agents

River AI has closed a $1.1 billion seed and Series A round just two months after emerging from stealth, according to TechCrunch. The startup, founded by Igor Babuschkin—a co-founder of xAI with prior roles at DeepMind and OpenAI—is pursuing a fundamentally different approach to AI development.

General Catalyst and AMP PBC led the round, with participation from Nvidia, AMD Ventures, Y Combinator, and Temasek. AMP PBC is a new AI-focused investment firm launched in 2026 by former Andreessen Horowitz general partner Anjney Midha.

A different vision for AI agents

River's mission diverges sharply from the prevailing industry trajectory. While most AI labs are building systems designed to replace human workers, Babuschkin wants to create personally trainable assistants that users control and customize themselves.

"Capable agents will be a normal part of everyday life. Less like the assistants you call on today when you need a task done, more like guardian angels: quietly present, on your side, helping with what actually matters to you," Babuschkin wrote in his June launch announcement.

Achieving this vision requires rebuilding the entire AI stack from the ground up, according to the company. That includes new approaches to training, model architecture, product design, and hardware that enables personal AI to run locally.

What River offers today

The company's first product is an API that allows developers to fine-tune open models using reinforcement learning and low-rank adaptation techniques. River bills usage per million tokens, with rates varying by the underlying open model selected.

The company positions this as an alternative to prompt engineering. Rather than steering models they don't own or control, developers can train open models into customized versions they fully possess and deploy like any standard API endpoint.

For enterprises, River promises a "neocloud" offering that eliminates the need for specialized post-training expertise. The company claims organizations can complete complex reinforcement learning runs in 15 to 20 minutes without dedicated infrastructure teams, at two to four times the cost savings compared to closed-source alternatives.

Why it matters

The enormous funding round for such a young company reflects both the frothy AI investment climate and a genuine shift in enterprise priorities. Organizations are increasingly seeking to control their AI destiny through open-weight models rather than depending entirely on proprietary systems from major vendors. River is betting that the missing piece—accessible, fast post-training capabilities—represents a significant market opportunity. The company's longer-term vision of individually trained personal agents aligns with emerging trends toward local AI execution, as evidenced by projects like OpenClaw and partnerships between Nvidia and PC manufacturers for AI-capable hardware.

Whether River's technical approach will deliver on its ambitious promises remains to be seen, but the company now has substantial resources to execute its vision. TechCrunch first reported the funding details.