Contribute your GPU, earn income.
Run a subnet node to serve AI models by offering your GPU for inference, and optimistically validate other nodes' inference. Nearly any GPU-enabled computer can contribute, and even a CPU-only server works for smaller models.
Mine from home
Earn income from home. Nearly any GPU-enabled computer can contribute to the network, no rig and no data center required.
No one in control
Like a mix of BitTorrent and Bitcoin mining, no single user controls the network and no one can control how users train and fine-tune their models.
AI on-chain
A blockchain designed for decentralizing, interfacing, incentivizing, validating, and democratizing decentralized AI models.
Fine-tune and infer
Employ any fine-tuning and sampling methods, run custom paths through a model, or read its hidden states, as simply as using PyTorch or Hugging Face.
Let's build.
Use any of our products or build your own, powered by the Liquid Tensor decentralized blockchain and its subnets.
Subnets
Decentralized AI model hosting, such as LLMs and image diffusors.
LEARN MORE →Liquid Tensor GPT
Chat with Hi, the Liquid Tensor GPT, running on decentralized AI.
LEARN MORE →HiGPTOnX
The Hi GPT, but on X.
COMING SOONSearch Engine
Search as if it's Google, powered by decentralized AI.
COMING SOONAPI
Customize your own application using the Liquid Tensor network.
COMING SOONCommunity Apps
Build your own applications and be featured.
COMING SOONHold $lTAO, run a node, mine lTAO.
Now you do not even need a GPU to take part. Hold $lTAO and a Subnet Node mints to your wallet, mining lTAO on a Bitcoin-style halving emission.
21,000,000 fixed supply: 20% seeds the Uniswap v4 pool, 80% mined to nodes over 64 yearly halvings by stake-weight. 600 $lTAO per node, 2,100 max buy and 42,000 max wallet, 4% swap fee to the treasury. Full mechanics in the brief.
Built by builders.
Kade Mauer
Distributed systems and cryptoeconomics. Previously built peer-to-peer compute infrastructure.
Sora Ellison
Decentralized model training and optimistic inference validation across subnets.
Devin Okoro
Uniswap v4 hooks, the emission engine, and the on-chain Subnet Node renderer.
Marlo Vance
Operator onboarding, partnerships, and the voice behind Liquid Tensor on X.