- Nosana (NOS) research overview
- Historical market behavior
- YearBull metric interpretation
- Market structure and supply
- Key risks and limits
- Primary sources and review scope
- Nosana: A Solana-Based Marketplace for AI GPU Compute
- What Nosana is designed to do
- How the protocol is organized
- The role of NOS
- Payments, staking, and governance
- Code, upgrades, and operational control
- What to verify before relying on the network
- Key takeaways
- Risks and open questions
- YearBull Rank update
Nosana (NOS) research overview
Nosana (NOS) is tracked by YearBull under the source identifier nosana. Source categories place the asset in the AI Cryptocurrencies universe, with additional labels including Artificial Intelligence (AI), Solana Ecosystem, DePIN. Category labels describe market context; they do not prove project activity, adoption, or investment quality.
Market structure and supply
Observed market capitalization is about $28.45 million and reported 24 hour volume is about $389.8 thousand. That volume equals 1.37% of market capitalization in the dated snapshot. Current circulating supply is 99,999,719. The recorded maximum supply is 100,000,000. Circulating supply changed +19.9% across the available historical window. Reported volume and supply fields can change through source revisions, issuance, burns, migrations, or venue coverage.
Key risks and limits
Liquidity depth, holder concentration, contract or network controls, token issuance, venue availability, governance, and operational dependencies remain material. High YearBull Risk appeared on 2.4% of stored observations. Historical metrics describe the available YearBull record; they do not predict future returns. Contract addresses, network support, custody, and venue availability should be verified before use.
Primary sources and review scope
YearBull methodology | Official project website | Source repository. Identity, categories, supply, and historical market fields were reviewed from locally stored source records on 2026-09-12. The live analytical snapshot may be newer than this editorial review.
Nosana: A Solana-Based Marketplace for AI GPU Compute
Nosana links people who need GPU capacity with independent hardware operators through Solana programs, a hosted deployment platform, and the NOS token. Its practical value depends on available hardware, workload demand, software reliability, and the controls governing protocol upgrades.
What Nosana is designed to do
Nosana is a compute marketplace for AI workloads. Its stated model connects clients that need GPU capacity with independent hosts that provide GPU machines, rather than relying on one centralized cloud operator. The project describes the network as a combination of Solana smart contracts, open-source node software, and hardware-level integration for GPU workloads. This is a project description, not independent evidence that the marketplace is cost-competitive or has reached a particular level of adoption.
The intended users fall into two broad groups. Developers and organizations can submit inference jobs or deploy AI applications, while GPU owners can connect hardware to the network and receive NOS-based rewards. The current documentation also describes a hosted dashboard, REST API, TypeScript SDK, and command-line interface, giving users several ways to submit or manage workloads.
How the protocol is organized
Nosana uses a set of Solana programs rather than one general-purpose contract. The project documentation identifies separate programs for staking, rewards, pools, and jobs. Together, these programs are intended to coordinate token staking, GPU-node registration, workload submission, reward distribution, and token pools. The documented program addresses provide a more precise technical reference than the broad DePIN label used in market classifications.
At the application layer, a workload is represented as a job or deployment. The documentation distinguishes individual jobs from deployments, which manage job lifecycles, and from GPU markets, where resources are organized for scheduling. The CLI repository shows that jobs can be posted as container workloads and that nodes can be started with Docker or Podman-compatible infrastructure. This creates a practical dependency on container images, host operating systems, networking, and the availability of compatible GPUs.
The role of NOS
NOS is the native Solana SPL token associated with the network. The official token documentation lists a fixed total supply of 100,000,000 NOS and identifies the token mint as nosXBVoaCTtYdLvKY6Csb4AC8JCdQKKAaWYtx2ZMoo7. The project describes NOS as usable for network compute and as part of the economic system for hosts, jobs, staking, rewards, and pools.
The token documentation divides supply into public-sale, airdrop, mining, team, liquidity, company, and backer pools. It records different release schedules for those pools, including stated vesting periods for mining, team, company, and backer allocations. Those schedules explain potential distribution pressure, but they do not by themselves establish current circulating supply or holder concentration; those questions require current on-chain analysis.
Payments, staking, and governance
Nosana’s protocol documentation says users can stake NOS to receive xNOS and participate in governance. It also describes staking and rewards as separate program components. However, the reviewed materials do not establish the full scope of governance authority, voting thresholds, quorum rules, or whether all material operational decisions are controlled by token holders. Readers should therefore treat governance participation as a documented feature, not as proof of fully decentralized control.
The platform also introduces an account and payment layer above the base token system. The getting-started documentation refers to prepaid credits for compute and supports dashboard, API-key, wallet, and command-line authentication paths. This suggests that using the product may involve platform accounts, credits, and service interfaces in addition to direct Solana transactions. The legal terms identify Nosana BV as the service provider and say the platform connects clients with independent hosts, creating a distinction between the company-operated service layer and the underlying protocol programs.
Code, upgrades, and operational control
Nosana publishes repositories for its programs, node software, CLI, dashboard, and related tools. The program repository states that the mainnet programs are deployed and that upgrades are prepared through continuous-integration processes and executed through a Squads multisig. That is a meaningful control point: even when program code is public, the ability to change deployed programs depends on the upgrade authority and the signers who control it.
The same program repository warns that most of the code is unaudited. This does not demonstrate a vulnerability, but it limits what can be concluded about security from public availability alone. Users also depend on the Solana network, RPC providers, container runtimes, GPU drivers, external model or container artifacts, and Nosana’s service interfaces. A failure in any of these layers could affect job execution even if the token contracts continue operating.
What to verify before relying on the network
The central open question is not whether Nosana has a token or published software, but how reliably the complete system serves useful workloads at acceptable cost. Prospective users would need to examine available GPU markets, queue behavior, job success rates, data-handling requirements, geographic distribution, and the terms attached to the hosted platform. Token holders would additionally need to monitor allocation releases, staking mechanics, upgrade-authority arrangements, and the relationship between NOS demand and any credit-based payment flow.
Key takeaways
- Nosana is structured as a Solana-based marketplace connecting GPU hosts with AI workload users.
- Its architecture separates staking, rewards, pools, and jobs into distinct Solana programs.
- NOS is used within the documented network economy and has a stated maximum supply of 100,000,000 tokens.
- The user experience includes dashboards, credits, APIs, SDKs, and a CLI, so the product is not limited to direct wallet interactions.
- Public code does not equal immutable or independently audited code; the project documents multisig-controlled upgrades and warns that most programs are unaudited.
- Real-world usefulness depends on GPU availability, job reliability, operating costs, and demand for the platform.
Risks and open questions
- The official program repository says most of the protocol code is unaudited, leaving unresolved smart-contract and implementation risk.
- Mainnet program upgrades are described as being executed through a Squads multisig, so users should assess upgrade authority, signer concentration, and operational procedures.
- GPU supply, queue performance, job completion rates, and geographic distribution are not established by the reviewed documentation.
- The platform depends on Solana, RPC infrastructure, container runtimes, GPU drivers, external artifacts, and Nosana-operated interfaces.
- Token allocation schedules may create distribution or selling pressure, while current holder concentration and circulating supply require separate on-chain verification.
- The reviewed materials describe governance participation through staking and xNOS but do not fully specify voting rules or the boundaries of token-holder control.
YearBull Rank update
YearBull Rank now for nosana: #5373.
Rank movement (time windows).
Reading rule: a smaller rank number indicates stronger placement.
- 7d window (2026-09-30): #1712 → #5373 (down by 3661).
- 30d window (2026-09-07): #2908 → #5373 (down by 2465).
Stability posture: the same move can be stable in one market and fragile in another.
Liquidity context: a quiet tape can still re-rank the pack.
Venue context: a broader footprint often smooths the rank trajectory.
Cycle read: a quick bounce can still be a mean-reversion phase.
YearBull Rank is a comparative ordering used on YearBull to place a coin versus others using a consistent set of inputs. Smaller numbers mean the coin sits higher in the YearBull list. It is meant for comparison and tracking, not certainty.

