River AI, the startup xAI co-founder Igor Babuschkin took out of stealth on June 10, 2026, has raised $1.1 billion across its Series Seed and Series A rounds, the company announced on August 11, 2026. General Catalyst and AMP PBC led the financing, with strategic investment from Nvidia and AMD Ventures and participation from Y Combinator and Temasek.
River is two months out of its launch post and already carries one of the largest early financings of the year. The money backs an unusual thesis: that the AI stack, from training through models, product, and hardware, needs to be rebuilt around personal agents that people own rather than intelligence rented from a handful of labs.
“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,” Babuschkin, River’s co-founder and CEO, said in the announcement.
What River Is Actually Selling First
The company’s first product, live since the June launch, is the River API, a training service for open-weight models. It offers LoRA fine-tuning and reinforcement learning on open models ranging from 35 billion to 1 trillion parameters, including Qwen3.6, Kimi K2.6, and GLM 5.2, with billing metered per million tokens for both training and inference rather than by GPU hour.
The pitch is aimed at enterprises that want models shaped around their own work without standing up an infrastructure team. River says any enterprise can complete a complex reinforcement learning run in 15 to 20 minutes with no infrastructure team required, at two to four times the cost savings relative to closed-source alternatives, and that trained checkpoints belong to the customer and deploy to an OpenAI-compatible endpoint.
The pricing page puts concrete numbers on the meter: training runs $1.00 per million tokens on the Qwen3.6 35B model and $12.84 per million on Kimi K2.6 at the 262k context length, with checkpoint storage billed separately at $0.10 per gigabyte per month. A worked example on the site prices a full production-scale RL run on a math dataset at under $1,000.
The Longer Bet Behind the Round
Babuschkin spent roughly a decade at DeepMind, OpenAI, and xAI, where he co-founded the latter and worked on large-scale training, before starting River. In his launch post, he framed the company against the direction the frontier labs are heading: agents that replace workers, rented out by the token.
“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. They will know you well, and they will be yours, not someone else’s,” he wrote.
That framing, personalization and continual learning on top of a training API, with personal hardware further out, is what General Catalyst is buying. “American leadership in AI urgently requires leadership in open-weight models, while maintaining a lead in closed frontier models. Igor and the River AI team have the experience to make this happen,” said Hemant Taneja, General Catalyst’s CEO, in River’s announcement. Marc Bhargava, the firm’s managing director, pointed to the commercial gap: companies have lacked a cost-efficient way to train, tune, and own custom models, and River “closes this gap.”
AMP PBC, the co-lead, joined General Catalyst in leading the round, per River’s announcement. Nvidia and AMD Ventures taking strategic stakes puts both GPU rivals inside the same cap table of a company whose stated roadmap ends in hardware that runs personal AI outside someone else’s data center.
Where the Money Lands
River says the funding accelerates every layer of its stack at once: the training API it has already shipped, a personalization and continual-learning product layer still being built on top of it, and new hardware intended to let personal AI run close to its owner. The founding team draws on xAI and Tesla engineering experience, and the company is hiring.
The structure of the round is worth noting on its own. A combined seed and Series A of this size, announced two months after a launch post, with strategic chip investors alongside financial leads, gives River the capital to compete for training compute and research talent against the frontier labs its founder came from. Whether the personal-agent thesis holds is a longer question; the immediate business is a post-training cloud for enterprises, and that business now has a billion-dollar runway and two venture firms behind it.
