River AI (river.ai), a Palo Alto-based company building full-stack infrastructure for personal AI owned by individuals rather than corporations, has raised $1.1 billion in combined Series Seed and Series A funding led by General Catalyst and AMP PBC, with participation from Nvidia, AMD Ventures, Y Combinator, and Temasek.
The capital will accelerate development of an API for LoRA-based fine-tuning and reinforcement learning on open-weight models ranging from 35B to 1T parameters, along with longer-term plans for personal hardware.
Open Models Shift Enterprise AI
The timing aligns with a broader move toward customized, user-owned AI. Together AI raised $800M Series C in July 2026, while Fireworks AI secured $1.5B Series D the same month. River differentiates by emphasizing reinforcement learning for personalization and full model ownership, rather than inference speed alone.
Enterprises Seek Model Ownership
Current closed API providers force companies to send proprietary data to external labs without retaining control or improving models over time. River's approach lets enterprises run complex RL training in 15-20 minutes at 2-4x lower cost than closed alternatives, according to the company.
Full-Stack Personal AI Stack
River is rebuilding the AI stack across hardware, infrastructure, models, and interfaces. Its API already supports pay-per-token fine-tuning and RL on open models, with a console at console.river.ai. This contrasts with inference-focused platforms like Together AI and Fireworks AI, which prioritize throughput over continual personalization.
"To get there, we believe the stack has to be rebuilt end to end: training, models, the product layer, and new hardware that lets personal AI live close to you." — Igor Babuschkin, CEO
Elite Backers Signal Strategic Bet
General Catalyst led the round alongside AMP PBC, founded by former a16z partner Anjney Midha. Strategic investments from Nvidia and AMD Ventures indicate hardware ambitions, while Temasek and Y Combinator add global and early-stage validation. CEO Hemant Taneja of General Catalyst framed the bet as advancing American leadership in open-weight models.
Open AI Infrastructure Market Expands
The enterprise AI market reached $42B in 2026, with fine-tuning and training segments growing at 18-24% CAGR. Competitors include Cohere ($1.54B raised) and Anthropic ($14.5B+), yet River's open-weight, ownership-focused thesis targets a distinct gap: enterprises tired of renting models they cannot own or continually improve.
xAI Co-Founder Leads Ambitious Effort
Igor Babuschkin previously co-founded xAI and held senior roles at OpenAI and DeepMind. The team is hiring across hardware engineering, research, and product to support the full-stack vision.