River AI Raises $1.1B at $5B Valuation to Let Enterprises Own Their AI Models

Quick Facts

River AI raised $1.1 billion across a seed and Series A at a roughly $5 billion post-money valuation.
Investors include General Catalyst, AMP PBC, Nvidia, AMD Ventures, Y Combinator, and Temasek.
The company’s API lets developers run reinforcement learning training runs in 15 to 20 minutes at two to four times lower cost than closed-source alternatives.

River AI Inc. has raised $1.1 billion in early-stage funding just two months after its public launch, making it one of the largest early financings of 2026. The deal values the company at roughly $5 billion.

General Catalyst and AMP PBC led the round, which was structured as convertible preferred shares. Nvidia Corp., AMD Ventures, Y Combinator, and Temasek also participated.

The startup was founded by Igor Babuschkin, who co-founded xAI Corp. and previously worked at Google DeepMind, where he helped build AlphaCode, the first coding AI to demonstrate competitive performance in a programming contest. Babuschkin departed xAI in August 2025 and incorporated River AI in Nevada in April 2026.

River AI’s founding thesis is direct: enterprises should own and customize their AI models rather than rent access to general-purpose systems from large labs. The company’s first product, the River API, is a cloud service that lets developers apply LoRA fine-tuning and reinforcement learning to open-weight large language models ranging from 35 billion to 1 trillion parameters.

According to the company, training runs that would typically require a dedicated infrastructure team can be completed in 15 to 20 minutes through its platform. Billing is metered strictly on tokens used, eliminating idle GPU costs. The platform manages weight transfers, sampling-training consistency, and elastic compute automatically.

Babuschkin described the broader goal in his launch blog: “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.”

General Catalyst CEO Hemant Taneja framed the investment in national terms. “American leadership in AI urgently requires leadership in open-weight models, while maintaining a lead in closed frontier models,” he said. “Igor and the River AI team have the experience to make this happen, and we view their agenda as a priority for American resilience.”

AMP PBC, the other lead investor, was founded in 2026 by Anjney Midha, a former general partner at Andreessen Horowitz who backed Black Forest Labs, Mistral AI, and OpenRouter while at a16z.

The River API is the first layer of a broader product suite. The company says its next release will focus on personalization and continual learning for AI agents. Beyond software, River AI is also developing custom silicon. A job posting confirms plans to build a system-on-chip with an onboard machine learning accelerator manufactured on advanced foundry nodes, paired with a compiler that converts PyTorch-based models for the chip.

The presence of both Nvidia and AMD Ventures on the same cap table is notable. The two GPU makers are direct competitors in AI training hardware, and River AI’s stated hardware roadmap could eventually position its chip as an alternative to both. The investments suggest each company sees value in backing the broader open-weight AI software ecosystem, regardless of which hardware it ultimately runs on.

For enterprise software buyers, River AI’s model represents a direct alternative to API-based access from OpenAI, Anthropic, and Google. If the company’s cost and speed claims hold at scale, it gives technical teams a path to owning fine-tuned models without standing up dedicated GPU infrastructure.

This article was written by an AI agent. Spotted an error? Send a correction and we will fix it.