Arbius is a decentralized network for machine learning, built on Arbitrum. Arbius enables censorship resistant AI models and agents by rewarding AI miners (Proof of AI Compute) who deliver provably verified AI output directly to application users It has a token with a fixed supply like Bitcoin.
The Arbius Protocol can be used to run any AI model such as image generation, video generation, tuning, large language models (LLM), and seamlessly integrate other downstream product into decentralized applications. By connecting miner (GPU supplier) and user directly, Arbius allows for a decentralized marketplace for AI models that generate fast results for a market competitive fee.
New coins are generated with GPU power by participating in the network. There is no central authority to create new coins, and holders vote on-chain for protocol upgrades. Arbius is fully open-source, and models operate as DAOS with custom rules for distribution and rewards, providing a way for model creators to earn income.
At the time of writing, Arbius offers two incentivized pool options through GYSR (v1 and v2 Arbius) — two separate competitive geysers for gamified staking. The first geyser locked up 18,000 AIUS tokens to be distributed over the course of a full year, working to align longer term incentives of stakers and the project.
We’re excited to follow Arbius’ growth over the coming months and years as the team continues their important work democratizing the future of AI. For more information about Arbius, check out their website and follow them on X.
”We launched with GYSR on Day 1 of our protocol, because we really liked how the incentives were designed, and the simple manner in which existing LP can be staked, providing an easy to use pathway for us to give rewards. We like that the system automatically adjusts incentive without centralized intervention. We look forward to working with them and for more users to discover GYSR.”
- J Master Pig, COO, Arbius.ai
—
Disclaimer: GYSR is not associated with any of the projects we profile in our spotlight series and these posts should not be seen as endorsements. We encourage readers to do their own research.
