# River AI Secures $1.1B Investment to Revolutionize Personal AI A major step towards launching the era of open-source personal artificial intelligence. - **Source:** [Unite.AI](https://www.unite.ai) - **Published:** 2026-08-11 - **Topic:** [Data & AI](https://brifing.si/c/data-ai) - **Original article:** - **On Brifing:** https://brifing.si/s/a8603e28-89c2-4a82-b1b4-59c8fe511b69 ## River AI's Grand Entrance River AI, co-founded by xAI's Igor Babuschkin, emerged from stealth in June 2026 and announced a $1.1 billion investment in August 2026. The Series Seed and Series A rounds were led by General Catalyst and AMP PBC, with participation from Nvidia, AMD Ventures, Y Combinator, and Temasek. This funding supports the vision of rebuilding the AI stack around personal agents that individuals can own. ## The Open-Source AI Vision River AI founder Igor Babuschkin argues that current AI systems are overly dependent on labs and need to change. Babuschkin advocates for AI that's open, accessible, and personalized. River AI aims to build a future where individuals and companies can own their intelligence. This vision has attracted significant investor support, marking a shift in how AI development might evolve. ## Training Models with River API River AI's first product, the River API, offers training services for open-weight models. It enables LoRA fine-tuning and reinforcement learning on models ranging from 35 billion to 1 trillion parameters, supporting models like Qwen3.6, Kimi K2.6, and GLM 5.2. The platform uses a per-million-token billing system rather than GPU-hour pricing, providing cost advantages for users who need scalable training solutions. ## Enterprise Advantages The River API promises to complete complex reinforcement learning runs in 15-20 minutes without requiring infrastructure teams. Offering 2-4 times cost savings compared to closed-source alternatives, the system guarantees that trained models belong to customers and can deploy to OpenAI-compatible endpoints. These features create an attractive solution for enterprise customers seeking customizable AI solutions. ## Pricing Structure and Example Costs River API's pricing varies by model and context length. Training costs $1.00 per million tokens on the Qwen3.6 35B model, rising to $12.84 on Kimi K2.6 at 262k context length. Checkpoint storage is billed separately at $0.10 per gigabyte per month. A worked example on their site shows a full production-scale RL run on a math dataset costing under $1,000, demonstrating competitive pricing for enterprise-scale operations. ## Founder's Background and Vision After nearly a decade at DeepMind, OpenAI, and xAI (which he co-founded), Igor Babuschkin launched River AI. Babuschkin envisions personal agents becoming a normal part of daily life - not like today's task assistants but more like guardian angels: quietly present, aligned with users' interests, and helping with what truly matters. This compelling vision attracted significant investor interest in River's alternative approach to AI development. ## Investor Perspectives General Catalyst CEO Hemant Taneja highlights River AI's potential to lead in open-weight models while maintaining closed-model advantages. Managing director Marc Bhargava notes that companies lacked cost-efficient ways to train, tune, and own custom models, a gap River fills. Strategic investments from GPU rivals Nvidia and AMD Ventures also support River's hardware roadmap for personal AI outside traditional data centers. ## Next Steps and Industry Impact With new funding, River AI plans to expand its training API, develop personalization and continual-learning product layers, and create hardware enabling personal AI to run close to its owner. Drawing on xAI and Tesla engineering talent, the company continues hiring. These moves could pioneer open-source, personalized AI solutions, potentially reshaping how artificial intelligence integrates into daily life and business operations. ## Sources covering this story - https://www.unite.ai/river-ai-raises-1-1b-out-of-stealth-to-rebuild-the-stack-for-personal-ai/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._