Skip to main content

Overview

ASI:Cloud is a permissionless GPU Cloud and AI inference platform developed by the ASI Alliance (CUDOS × SingularityNET). It evolves CUDOS Intercloud into the AI compute layer for the ASI Alliance and the wider AI developer ecosystem, prioritizing open access, transparent pricing and per-pay-token access to state-of-the-art open-source Large Language Models (LLMs).

At a glance:

  • GPU Cloud: Per-hour VM/GPU rentals, private networks and pre-set VM templates.
  • Serverless Inference: OpenAI-compatible endpoint, one-click model selection, multi-API-key management and a model chat playground.
  • Developer experience: OpenAI API parity for quick integration, usage dashboards and competitive, transparent pricing.
  • Web3-first access: Wallet login (Metamask / ASI / Kepler), permissionless (no KYC). Crypto payments today; fiat support coming soon.
  • Developed and operated by experts: Inference service is provided by SingularityNET, an AI R&D powerhouse. GPU infrastructure operated by CUDOS (the web3 arm of CUDO) with decades of cloud computing and infrastructure experience.
note

Release note: The initial September 2025 release focuses on the beta of the LLM inference service as well as GPU Cloud features previously available via CUDOS Intercloud. Future roadmap includes support for additional AI workloads, multimodal inference (speech/image/video) with ongoing performance and model catalog updates and encrypted knowledge graphs.