Enable peer-to-peer distribution to scale your AI workloads across multiple devices. Share instances, shard models, and pool computational resources across your network.
Load balance across multiple instances
Split large models across workers
Pool resources from multiple devices
Start LocalAI with P2P enabled
local-ai run --p2p
This will automatically generate a network token for you.
Or use an existing token
export TOKEN="your-token-here"
local-ai run --p2p
If you already have a token from another instance, you can reuse it.
Access the P2P dashboard
Once enabled, refresh this page to see your network token and start connecting nodes.