decentralized-compute

A decentralized compute network that enables devices to perform computational work for one another in exchange for value.

Decentralized compute protocol turning idle devices into verifiable AI infrastructure. Mobile-first, open, and accessible.

Proof-of-Contribution Chain: traceable contributions in a Web3-adjacent ecosystem (without classic blockchain complexity) earn voting power on roadmap priorities. Closed Gem Economy: gems act as currency in the NexRealm Marketplace for assets, earned through real contributions. Cross-Project Gems: gems become portable across creative ecosystems.

Miraset Chain: A decentralized GPU compute marketplace for AI inference, powered by Proof of Compute & Capacity (PoCC) and built in Rust.

Citrate L1 — AI-native BlockDAG blockchain: GhostDAG consensus, EVM/LVM execution, node, CLI, contracts, ZK circuits (chain 40204).

DeCub (Decentralized Compute Platform) is a decentralized distributed compute and storage platform using a hybrid consensus model (local RAFT + global Byzantine Fault Tolerance) plus gossip, object storage, and cryptographic proofs — designed for secure, scalable workload execution across clusters. REChain-Network-Solutions — focused on decentrals.

Prime Intellect — independent third-party profile of a public API surface, by API Evangelist. Prime Intellect is a San Francisco–based startup building an open and decentralized stack for developing and improving AI agents through reinforcement learning. The company combines a 50+ provider GPU compute marketplace (single-node and 1–256 GPU multi-no

A verified, bilingual state of the art of decentralized AI compute (training + inference) outside Google/Amazon/Microsoft datacenters. What is real, what is hype.

BasedAI is a proposed decentralized network for running LLM inference

matrix-compute-gate is an autonomous local transaction proxy gating VRAM compute slots. It manages an adversarial 10-day flat-file ledger where AI nodes settle satoshi micro-payments over a Tailscale mesh tunnel, calculating non-linear architectural transitions and data fortresses entirely offline in RAM.

Citrate compute pool — coordinator + workers for pooled AI training and inference (Groves).

Rent GPUs by the hour on Lium to top LocalMaxxing leaderboards with verified, high-speed inference for under $10 total.

Rust reference node for the Decentralized Universal Compute Protocol (DUCP).

Open-source decentralized hosting & compute platform. Deploy apps via Git or CLI across independent nodes with Docker, Traefik, FastAPI, distributed scheduling, rollback, and optional Solana devnet node-operator credits.

GRID peer-to-peer useful-compute network: Proof of Resource, blockchain, wallets, mining, and isolated container execution.

Wombo — independent third-party profile of a public API surface, by API Evangelist. Wombo is the AI company behind Dream by WOMBO, a consumer AI art and video generator, and w.ai, a decentralized AI supercomputer that pools idle GPU compute worldwide to serve affordable on-demand inference.

Mawari decentralized compute network node setup and configuration

Reproducible GPU inference benchmarks and DID-signed compute evidence

Turn your GPUs into an AI inference provider on Swan Chain. Starts and manages llama.cpp, vLLM and SGLang model servers, connects outbound over WebSocket (no inbound ports, no wallet), and includes a local OpenAI-compatible gateway and a monitoring dashboard.