List of Compute Network projects on Solana (2026)

Dispersed is a decentralized GPU compute network providing builders, creators, and enterprises with on-demand access to distributed graphics processing capacity. The platform aggregates geographically dispersed GPUs operated independently worldwide, routing AI workloads including video generation, image processing, reinforcement training, inferencing, and agentic AI development. Users can select GPU models, performance parameters, and pricing that match their requirements while avoiding vendor lock-in.

A decentralized serverless cloud designed to address challenges in cloud computing, such as centralized trust, fragmented ecosystems, and privacy concerns.

Decentralized cloud platform offering computing, storage, and indexing services.

Cherry Servers Cloud Platform provides on-demand bare metal dedicated servers, virtual machines, GPU accelerators, block storage, and networking services through a unified API and web portal. The platform supports custom hardware configurations with AMD EPYC and Ryzen processors, NVIDIA GPUs, NVMe storage, and multi-region deployment across Europe, North America, and Asia. Users access infrastructure management through RESTful API, CLI tools, Terraform, and Ansible integrations with hourly or fixed-term billing and cryptocurrency payment options.

Crunch Protocol is a decentralized coordination layer built on Solana that distributes machine learning workloads across a global network of data scientists. The protocol enables Coordinators to create prediction challenges with defined rules and rewards while data scientists compete to build predictive models. On-chain smart contracts handle verifiable scoring and automated reward distribution in USDC, with model and data privacy protected through Trusted Execution Environments.

Leea Protocol implements secure decentralized infrastructure for AI execution through encrypted virtualization, enabling private agent deployment and inter-agent communication. The system uses remote attestation and AMD SEV/Intel TDX for zero-trust security while maintaining blockchain-based node accountability.

Compute Labs provides a unified solution that tokenizes enterprise-grade GPUs into GNFTs, enabling trading, staking, and the creation of AI-Fi financial instruments like ETFs and perpetual contracts.

Arcium Network implements decentralized confidential computing through distributed Multi-Party Computation infrastructure, enabling encrypted computation across blockchain and off-chain applications. The system coordinates distributed Arx nodes that form clusters to execute secure computations defined within Multiparty eXecution Environments. Each node processes encrypted data fragments without accessing complete information, maintaining confidentiality through cryptographic protocols including MPC, homomorphic encryption, and zero-knowledge proofs. The architecture supports parallel execution of independent computations while orchestrating node coordination, resource allocation, and verification through blockchain integration.

Metal offers high-performance bare metal servers that can be deployed globally with low latency. Ideal for processor-intensive workloads, it provides dedicated hardware for applications in gaming, streaming, and Web3.

Accelerate provides powerful GPU instances tailored for demanding machine learning and AI workloads. It supports training, fine-tuning, and inference tasks, ensuring optimal performance for computationally intensive applications.

AIxBlock Platform provides end-to-end AI development infrastructure through decentralized computing resources, enabling model training, fine-tuning, and deployment. The system supports MCP integration and offers automated workflow creation for AI applications.

Inference API is a serverless inference tool for running AI models. It helps users process language and vision tasks through a global GPU network.

Raiku Coordination Layer implements high-performance block building and advanced scheduling for Solana, facilitating guaranteed transaction inclusion through validator coordination. The system provides sub-50ms pre-confirmations, ahead-of-time blockspace reservations, and deterministic execution through its Coordination Engine, Validator Sidecar, Slot Marketplace, and Global Accounts infrastructure. The architecture enables developers to integrate with as little as two lines of code while maintaining seamless composability with Solana's security model

Bless Network provides decentralized edge computing through dynamic resource matching and WASM secure runtime, enabling on-demand CPU and GPU allocation across distributed devices. The system orchestrates computational tasks while maintaining security through randomized distribution and modular verification mechanisms.

Expert-driven platform delivering multimodal training data through configurable workflows for data collection, annotation, QA, and human evaluation using domain experts.