llms.blog
AI

River AI Secures .1B Led by General Catalyst to Build Open-Weight Model Infrastructure

River AI, an artificial intelligence startup founded by former xAI co-founder Igor Babuschkin, has secured $1.1 billion in early-stage funding to build an open-weight model stack and decentralized AI infrastructure platform. The financing round was led jointly by General Catalyst and public benefit corporation AMP PBC, with strategic participation from NVIDIA, AMD Ventures, Y Combinator, and Singapore sovereign fund Temasek. Babuschkin, whose prior engineering background spans OpenAI, Google De

axc
axc
2 min
River AI Secures .1B Led by General Catalyst to Build Open-Weight Model Infrastructure

River AI, an artificial intelligence startup founded by former xAI co-founder Igor Babuschkin, has secured $1.1 billion in early-stage funding to build an open-weight model stack and decentralized AI infrastructure platform.

The financing round was led jointly by General Catalyst and public benefit corporation AMP PBC, with strategic participation from NVIDIA, AMD Ventures, Y Combinator, and Singapore sovereign fund Temasek. Babuschkin, whose prior engineering background spans OpenAI, Google DeepMind, and Tesla, is also committing up to $100 million of personal capital into the venture. The funding values the company at up to $5 billion.

River AI Modular Infrastructure

Decentralized AI Infrastructure and Turnkey Fine-Tuning

Unlike frontier AI labs focused primarily on monolithic closed-source models and enterprise labor automation, River AI is designing an end-to-end platform tailored for user-owned and customizable intelligence.

The company's core platform aims to solve operational bottlenecks for developers and enterprises seeking to fine-tune and serve open-weight models. Key technical and infrastructure elements include:

  • Parameter-Efficient Adaptation at Scale: Turnkey APIs providing Low-Rank Adaptation (LoRA) and reinforcement learning fine-tuning directly on frontier open-weight foundation models.
  • Token-Metered Deployment: Elimination of rigid reserved-cluster contracts, allowing teams to deploy tuned models to production with usage-based billing.
  • Infrastructure Abstraction: Automated orchestration that manages low-level kernel compilation, distributed serving schedules, and memory sharding without requiring dedicated in-house infrastructure engineering teams.

Shifting Focus to Personal and Open Weights

Babuschkin stated that the current trajectory of closed AI development centralizes control within a handful of model providers. River AI is structured around the thesis that intelligence should be open, portable, and aligned with individual user directives rather than provider constraints.

The substantial participation of hardware vendors NVIDIA and AMD Ventures highlights increasing compute demand for hybrid deployment ecosystems, where enterprises balance frontier closed APIs with locally optimized open-source weights.

Sources

AIInfrastructureOpen SourceFunding
ShareX (Twitter)LinkedIn

Written by

axc

More to read

  • Meta Prepares Consumer AI Agent 'Hatch' and October Launch for 'Watermelon' Frontier Model

    Meta Platforms is preparing to roll out an autonomous consumer AI agent codenamed Hatch in late August or early September, followed by the planned release of its next flagship foundation model, codenamed Watermelon, in October 2026. The initiatives, first reported by The Information, highlight Meta's dual-track approach to commercialize autonomous software workflows while scaling foundation model training compute to compete directly with frontier offerings from OpenAI and Anthropic. Consumer

    1 min

  • Continuous LLM Performance Profiling in Production: Roofline Models, Model FLOPs Utilization, Model Bandwidth Utilization, and Hardware Bottleneck Diagnostics

    Evaluating the runtime performance of large language model serving infrastructures requires looking beyond raw GPU metrics. Standard operating system utilities such as nvidia-smi report high GPU utilization percentages whenever compute cores or memory controllers are active, masking critical inefficiencies in memory access, communication, and kernel scheduling. A serving node running single-stream autoregressive decoding can report 100% GPU utilization while operating at less than 2% of the hard

    1 min

  • Latent Reasoning in Large Language Models: How Continuous Thoughts and Recurrent Hidden States Bypass Discrete Tokenization

    Standard autoregressive language models solve multi-step reasoning tasks by generating explicit verbal scratchpads. Under the Chain-of-Thought (CoT) paradigm formalized by Wei et al. (2022), a Transformer expands its effective computational depth by emitting intermediate natural language tokens into the prompt context. Each emitted token provides an additional forward pass through the network's layers, transforming reasoning into a sequence of left-to-right text predictions. While language-base

    1 min