News
DealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory’s NVIDIA & AMD Enterprise ExpansionDealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory’s NVIDIA & AMD Enterprise Expansion
EveryNeocloud
Nosana

Nosana

Decentralised marketplace for on-demand GPU compute

Nosana is a peer-to-peer compute marketplace that connects users with a global network of GPU providers. It offers on-demand access to graphics processing units for artificial intelligence workloads, aiming to reduce costs and avoid vendor lock-in through open-source infrastructure.

Share on WhatsApp

  • Peer-to-peer marketplace for on-demand GPU access
  • Network comprises approximately 2,000 active nodes
  • Claims up to 6x lower costs than traditional clouds
  • Supports Nvidia 30, 40, 50 series and enterprise GPUs
  • Uses $NOS token for network coordination and staking

What it is

Nosana operates as a distributed compute network that transforms idle graphics processing unit power into an on-demand resource. It functions as a peer-to-peer marketplace, linking developers and researchers with a worldwide array of hardware providers. The platform supports artificial intelligence tasks, including training, fine-tuning, and inference, alongside rendering and simulation. By leveraging a global grid of approximately 2,000 active nodes, Nosana allows users to scale workloads without long-term commitments. The service emphasises cost efficiency, claiming savings of up to six times compared to traditional cloud providers, while maintaining workload isolation and security.

How it came about

The provided material does not contain specific details regarding the founding date, initial funding rounds, or historical pivots of the organisation. It describes the current team as globally distributed, with members possessing experience in cryptography, corporate leadership, and artificial intelligence. One team member studied AI at Amsterdam University. The company maintains a presence in the Solana ecosystem, evidenced by its participation in events such as Solana Breakpoint London 2026. The narrative focuses on its ongoing mission to make compute resources a public utility rather than detailing its corporate origins.

The compute

The network relies on fully open-source infrastructure to ensure transparency and prevent vendor lock-in. It supports various Nvidia GPU series, including the 30, 40, and 50 series, as well as enterprise models like the Nvidia Pro 6000 Server Edition. Users can deploy workloads via a dashboard or integrate with tools such as Codex, Claude, and VS Code through the Nosana MCP Connect feature. The platform utilises a native token, $NOS, to coordinate and secure the network, allowing participants to stake tokens. Workloads range from agentic AI tasks to high-performance rendering, managed through a software development kit that enables coding agents to monitor deployments.

Where the GPUs are

Nosana does not operate centralised data centres in the traditional sense. Instead, it utilises a globally distributed network of GPU providers. This decentralised approach allows compute jobs to run near end-users, potentially reducing latency for real-time applications. The infrastructure is described as a grid comprising approximately 2,000 nodes with varied hardware configurations. These nodes are actively running AI models across different regions, providing a flexible and scalable resource pool that grows as more providers share their idle capacity.

Customers and deals

Several organisations utilise the Nosana network for production workloads. AI Builders uses the platform for hands-on workshops, allowing participants to run models in real time. Alio, an AI security lab, employs Nosana for training risk detection models and conducting red-teaming simulations. Inferia leverages the network for reliable production-grade inference without managing physical infrastructure. OpenGPU uses it as a task-based marketplace for faster access to compute. Sogni, a decentralised AI art platform, reports reducing infrastructure costs by up to 60 per cent while serving thousands of image generations daily.

Money and ownership

It highlights the economic model of the network, where users pay only for the compute they use, often at lower rates than conventional clouds. Providers earn income by sharing their idle GPU capacity. The network is powered and coordinated by the $NOS token, which also plays a role in staking mechanisms to secure the system. The company positions itself as a cost-effective alternative, with claims of significant savings for users depending on the workload type.

Its place in AI

Nosana positions itself within the broader movement towards decentralised physical infrastructure networks. It aims to redefine access to innovation by treating compute as a public utility. The project is active in the Web3 and Solana communities, participating in industry events to engage with developers and partners. By offering an open-source alternative to proprietary cloud services, it seeks to empower builders with flexibility and transparency. Its role involves facilitating a free-market model for GPU access, challenging the dominance of centralised hyperscale providers in the AI compute sector.

Questions people ask about Nosana

What type of hardware does Nosana support?

Nosana supports a variety of Nvidia GPUs, including the 30, 40, and 50 series, as well as enterprise-grade cards like the Nvidia Pro 6000 Server Edition. The network consists of approximately 2,000 nodes with diverse hardware configurations provided by individual participants.

How does Nosana reduce costs for users?

The platform connects users with providers who have idle GPU capacity, creating a competitive marketplace. Nosana claims this model can offer savings of up to six times compared to traditional cloud providers, as users pay only for the specific compute resources they consume.

Is Nosana centralised or decentralised?

Nosana is a decentralised, peer-to-peer network. It operates on fully open-source infrastructure and relies on a global distribution of GPU providers rather than centralised data centres. The network is coordinated and secured using the $NOS token and staking mechanisms.

What kinds of workloads can run on Nosana?

The platform supports high-performance workloads including artificial intelligence training, fine-tuning, and inference. It is also suitable for rendering, simulation, and gaming tasks. Users can deploy jobs directly from a dashboard or integrate with developer tools like VS Code.

Good reading and sources

    Where this page comes from

    There is no English Wikipedia article on Nosana yet, so the facts on this page come from the sources listed above: nosana.com/about, nosana.com/. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryNeocloud wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.

    It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

    The weekly digest

    Every Monday: the provider of the day, new pages on EveryNeocloud, and the conferences and anniversaries of the week ahead.