TLDR: DePIN revenues and token values for decentralized compute networks fell sharply in 2025, and we are comparing Flux to our competitors to assess the current state of compute networks.
WHO: This article is for those considering app deployment across DePIN networks and seeking guidance on which network to choose.
NEXT STEPS: Review FluxCloud resource stats to gain a clear picture of Flux’s superior compute capacity when compared to other DePIN projects.
WHAT TO DO: Take advantage of FluxCloud’s Deploy with Git feature to seamlessly deploy an app directly from a Git repository and test our network’s reliability.
Introduction
Messari’s State of DePIN Report 2025 found that infrastructure revenues and token values are in the red and will continue to decline across every major project in the decentralized physical infrastructure network narrative.
Flux operates in the general-purpose compute sector of the DePIN landscape, with FluxCloud, a decentralized cloud network, and FluxEdge, a P2P distributed compute and GPU network, supplying processing power for app development and AI inference.
We know the Flux community may be struggling right now due to broader crypto market conditions. So, in today’s blog, we will provide a general overview of Flux’s current standing relative to other top general-purpose compute projects that also offer distributed processing power for app development. Let’s dive in!
Akash
Akash Network is a decentralized cloud marketplace for containerized applications, where development teams can purchase distributed compute resources tailored to their application specifications.
Users describe the resources they need (CPU/RAM/storage/GPU) in a YAML file, then independent compute providers bid on the file to host the app. Compute capacity on Akash is rented on short-term leases, and users can select hosting bids from providers based on resource pricing and location.
Once a provider is selected and the compute is leased, app containers run on the provider’s hardware, and users pay as they go.
Akash Network Resources:
•CPU (vCPU/cores): 9,286.80
•GPU (count): 265
•RAM: 76.28 TB
•Storage: 690.17 TB
Why Flux Over Akash?
Akash is a lease-driven cloud marketplace, meaning that compute providers come and go, leases expire, and deployed apps are guaranteed to remain reachable at the same endpoint. Users are responsible for managing their own endpoints if they have multiple deployments, their app requires external routing, or they need failover.
FluxCloud’s advantage over Akash is that it’s less about “compute supply” and more about “platform supply,” providing holistic runtime environments for deployments that ensure apps run stably and efficiently over time.
FluxCloud supports single- and multi-instance deployments, keeping application state consistent across multiple independent nodes. If one node fails, FluxCloud assigns runtime workloads to another instance, enabling automatic failover for greater network redundancy.
Now, Akash deployments are Docker-based: users tediously construct a Dockerfile → push it to an app framework registry → and then manually maintain states in an app manifest. FluxCloud’s Deploy with Git feature eliminates this process entirely: application frameworks are detected, and runtime environments are configured automatically, allowing users to deploy apps directly from GitHub repositories with just a repo URL and no Dockerfiles. Deploy with Git means apps will run without having to painstakingly register an app or stand up a container.
CUDOS
The CUDOS Intercloud is a globally decentralized cloud infrastructure that connects developers to scalable, cost-effective distributed resources for running high-performance computing (HPC) servers.
Users select a location and their desired GPU, and Intercloud spins up a virtual machine (VM) that serves as a standard server for content delivery and handling requests. Redundancy is achieved by spinning up and provisioning multiple VMs, and users must configure their own load balancing (distributing network traffic across multiple servers) to ensure data availability.
CUDOS Network Resources:
•CPU: 12,000+ processor cores
•RAM: 26,000 GiB
•Storage: 575 TB
Why Flux Over CUDOS?
CUDOS primarily provides servers that deliver static content for processed app data and API requests. Flux’s advantage is that it provides a broader range of services —not just servers but also runtimes (application execution environments) that execute logic.
CUDOS Intercloud VMs require users to manage their own application layers, including OS installation, patching, bug fixes, load scaling, and codebase audits. FluxCloud handles all of that with fully managed deployments, automatically balancing loads and pushing updates and new features.
Stratos
Stratos is a decentralized storage network that splits files and documents into encrypted shards and stores them across a distributed network of nodes. Stratos provides storage capacity and bandwidth for data transfers between nodes, enabling high availability for users.
Stratos Network Resources:
•Storage Capacity: 34.78 PB
•Storage Nodes: 1,538
•Total bandwidth: 150 GB/s
Why Flux Over Stratos?
Flux’s advantage over Stratos is its complete compute ecosystem, with standardized resource provisioning and extensive hosting capabilities. Whereas Stratos functions exclusively as a storage network, providing capacity, storage nodes, and bandwidth for file transfers.
Stratos is well-suited for decentralized storage and data delivery. Still, if you need an always-on infrastructure for compute scheduling, instance replication for failover, and stable service endpoints, FluxCloud is the right choice.
Golem
Golem Network functions similarly to FluxEdge in that it’s a decentralized compute marketplace where independent contributors provide resources to the network, which users pay to run app workloads on. Golem is an open-source protocol that allows anyone to join the network as a contributor. Contributors rent out excess compute and earn $GLM tokens for doing so.
Golem Network Resources:
•Contributors: 1,476
•CPU: 10,209 cores (14,521 threads)
Why Flux Over Golem?
Golem is built around execute-and-return workloads, in which tasks are submitted to the network in batches for bulk execution. These are short-term workloads that run until completion and involve little human interaction. Golem is ideal for temporary compute tasks, such as pushing an application update, but not for long-term workloads, such as app hosting.
Flux’s advantage over Golem is that it’s built around long-term, host-and-serve workloads that run continuously and wait for and respond to external API requests. Host-and-serve workloads are needed to run uninterrupted application hosting.
Additionally, Golem workflows are Docker-centric and require users to convert application Docker images into Golem Virtual Machine Images. Docker images are blueprints for applications that define all application logic and outline the app’s codebase. Golem offers image packaging that automatically converts Docker images, but users still must configure their own files first. F
FluxCloud’s Deploy with Git feature completely bypasses the need for Dockerfiles, allowing users to deploy apps directly from Git repositories. Simply paste a Git URL, click deploy, and let FluxCloud handle your app’s runtime, with new updates and changes to the app repo syncing automatically. No Docker images required.
Aethir
Aethir is a decentralized compute network designed for GPU-heavy workloads around AI training and gaming. Rather than offering general-purpose compute for typical daily app workloads, Aethir provides a large pool of distributed GPU resources for complex tasks.
Aethir Network Resources:
•GPU containers: 440,000+
•Geographic footprint: 94 countries, 200+ locations
•Cumulative compute delivered: 1.2B+ compute hours
Why Flux Over Aethir?
Flux’s advantage over Aethir is architectural: Aethir focuses on supplying raw GPU power, whereas Flux provides GPU power and the CPU-side components, such as state monitoring and an API layer, that apps need to function.
Most app deployments require more than compute to work; they also require background services that always run and demand continuous resource support, such as:
•The API that receives requests
•User login/authentication
•Rate limiting (to prevent abuse)Job queues
•Orchestration (coordinating everything)
•State MonitoringDatabases
FluxCloud is built to run this always-on backend service layer, keeping it accessible with automatic failover even when GPUs fail. Flux runs these always-on backend services as replicated instances, so when the primary node executing an app workload goes down, instances are automatically replaced, and workloads are reassigned without restarting.
Aethir excels at providing GPUs, but dev teams can underestimate the critical role of non-GPU infrastructure in app deployments. That’s where Flux takes the lead: it provides complete, resilient systems, GPUs, and all the accompanying backend app services.
Conclusion
If there is one takeaway from the state of general-purpose compute networks in 2026, it’s that decentralized resources support very different app runtime models. Akash operates a compute marketplace where capacity is supplied via leased machines.
CUDOS is akin to a traditional VM-based cloud network, where redundancy is built through user-managed instances, and scale is achieved through manual infrastructure tooling. Stratos is a decentralized storage network; Golem provides compute via a distributed GPU infrastructure; and Aethir supplies raw GPU power for complex AI and gaming tasks.
Compared with Flux, these other compute networks lack holistic runtime capabilities. Whether users are running a temporary workload, an “always-on” backend service that requires continuous compute, a storage container, or a full backend-to-frontend GPU pipeline for app deployment, Flux can deliver.
FluxCloud’s multi-instance hosting and automatic replacement of failed instances are designed to keep long-lived services reachable, maximizing redundancy. Additionally, Deploy with Git enables app deployments directly from Git repositories without requiring manually configured Dockerfiles.
Our network isn’t just a place to buy computing resources; it’s an entire runtime ecosystem for holistic app hosting and operations, from development to deployment. The future runs on Flux.