Run inference
Connect your tools to an OpenAI-compatible API and run prompts on models shared across the network.
Distributed AI inference
Big models.Shared power.
AI doesn't have to fit on one GPU. AIBnB splits models across everyday GPUs and Macs, bringing spare compute together to run your prompts.
macOS / Windows / Linux · Beta software
One app. Two ways to take part.
Use compute. Contribute compute.
Share a GPU
Have spare capacity? Open the app and turn on Contribute compute to help run the network.