River AI

River AI lets organizations fine-tune open-source AI models on their own data and task-specific examples, producing specialized models that are cheaper to run, more accurate for the specific job, and owned outright rather than rented from a vendor. It supports LoRA fine-tuning and reinforcement learning, ships OpenAI-compatible endpoints so there is no vendor lock-in, and prices by token usage. The company frames its mission against "the current landscape where the most powerful AIs are controlled by a handful of large corporations," aiming instead for personal AI that is "responsible, curious, and truly yours." It was founded by Igor Babuschkin (CEO), who spent nearly a decade building AI at DeepMind, OpenAI, and xAI before co-founding River, alongside Aaron Rogers, Bowen W, Dmytro S, and Yaroslav Nazarov. River AI raised a combined Seed and Series A totaling $1.1B, announced August 2026 and led by General Catalyst — an unusually large round for a company that had not yet publicly detailed customers or partnerships at the time of the raise.