When we launched Bless, our vision was clear but ambitious: to create a shared, decentralized compute network – a permissionless, community-driven alternative to the centralized cloud services we rely on today. Our goal was simple: allow anyone to contribute unused computing resources and give developers a platform to build on them instantly and affordably, at a global scale.
In just a few months, the network has surpassed even our most optimistic projections. But this is only the beginning. Today, we’re excited to share the roadmap that will guide Bless from a fast-growing testnet into a powerful foundation for the next generation of applications – ones that are AI-native, resilient, and open to all. Some details may evolve as we progress, but our mission remains unchanged: to shift computing away from centralized, corporate-controlled infrastructure and toward something shared, intelligent, and governed by the very builders and users who power it.
Laying the Foundation
We’re currently in the testnet phase, where we’ve laid the groundwork for Bless’s future. In November 2024, we launched the Bless Extension Node, enabling anyone with a browser to contribute idle compute power to a globally distributed network. Since then, Bless has grown rapidly, with over 5 million nodes launched and 1 million active daily nodes, spanning major regions around the world. Together, these nodes now provide more than 200 petaFLOPs of computing power – an impressive achievement for an open, permissionless network.
We also introduced the Bless CLI, which allows developers to deploy applications directly onto the network. This has sparked a growing ecosystem of web tools, automations, games, and experiments, demonstrating the potential of open and accessible compute resources. Soon after, we enabled external API calls, allowing hybrid applications that can interact with services and data outside the Bless network, opening new possibilities for real-world use cases.
Looking forward, we are focused on the AI-native future of computing. This isn’t just about AI applications, but a larger shift where AI becomes the dominant interface for how we write, trigger, and consume compute. Traditional request-response paradigms are giving way to autonomous agents and real-time inference loops, where AI models guide logic and user interactions.
To prepare for this future, we’ve integrated with ElizaOS, a framework for building AI agents. Developers can now run autonomous agents directly on Bless, using our network as the decentralized compute backend. This is a crucial step toward supporting intelligent, persistent agent networks.
We also launched Genesis Build, an initiative to reward the most creative projects built on Bless. More than a contest, this is a glimpse into the future we’re creating together, showcasing the possibilities when compute is abundant and accessible.
Scaling Capability and Integrity
Phase 1 of our roadmap marks a key transition as Bless evolves from early-stage success into a platform capable of handling more complex, AI-driven workloads. We’re introducing Bless Desktop Nodes, a more powerful version of our Extension Node. These nodes will support a broader range of workloads, including GPU resource sharing, enabling real-time inference, parallel processing, and heavier compute tasks. With these nodes, Bless will transition from a lightweight tool to a true execution backbone for the next wave of AI applications.
In this phase, we’ll also launch an anti-sybil campaign to ensure that rewards go to meaningful contributors – those who create real value for the network. This initiative will help preserve the integrity of the network as we scale, distinguishing true supporters from opportunistic testnet actors.
Phase 1 culminates with the mainnet launch, marking the transition of Bless into a viable, community-powered alternative to traditional cloud infrastructure. With scalability and developer-friendliness at its core, Bless will become a foundational layer for intelligent applications and composable software.
Maturing the Developer Experience
After the mainnet launch, Phase 2 will focus on enhancing the developer experience and expanding the capabilities of Bless. We plan to streamline deployment, allowing developers to deploy pre-packaged workloads (like Docker images) directly to the network. This will make it easier for Web2 and enterprise developers to move existing services onto Bless, with minimal reconfiguration.
As our orchestration layer matures, we’ll introduce Advanced Automated Scaling, allowing Bless to automatically allocate additional compute nodes to applications as demand grows. This will create a Web2-like developer experience, but with the added benefits of decentralized infrastructure.
We also plan to integrate WebGPU, allowing users of modern browsers to contribute GPU cycles, significantly increasing the range of workloads that can run on Bless – from real-time inference to graphics-heavy applications. This will lower the barrier to entry for developers while increasing the power of the network.
Phase 2 is about lowering friction while raising power – making it easier for developers to build serious applications without worrying about the underlying infrastructure.
Scaling Access and Incentives
In Phase 3, we’ll focus on scaling adoption, making it easier for developers to pay for services and ensuring that contributors are rewarded based on the real value they provide. We plan to introduce fiat and credit card payments, allowing users to pay for compute resources with traditional methods. This will make Bless more accessible to a wider audience, especially outside of Web3.
In this phase, we’ll also introduce Dynamic Rewards, a system that ensures contributors are compensated based on the quality, availability, and usage of their resources. Instead of fixed incentives, rewards will reflect the real economic value a node creates for the network. This approach will create a sustainable, self-balancing ecosystem, where great builders can easily create and great contributors are rewarded for their contributions.
Looking Ahead
From 5 million nodes to AI-driven applications, from credit card payments to dynamic rewards – our roadmap represents a new vision for compute, one where infrastructure is owned by the many, not the few. Each phase takes us closer to a platform that scales with ambition, adapts to developers’ tools, and invites anyone, anywhere, to participate.
The foundation is already in place. The network is growing. And the future is being written – one node, one agent, and one builder at a time.
