Inference Partners FAQ
This FAQ is for organizations interested in providing inference compute to the Public AI Gateway — our unified access point for open-source AI models.
About the Partnership Program
What is the Public AI Gateway?
The Public AI Gateway is a unified API that routes requests to open-source models hosted by inference partners worldwide. We handle load balancing, traffic routing, and failover so developers can access public AI models through a single OpenAI-compatible endpoint.
The gateway currently hosts models from partners including Swiss AI, AISingapore, Allen AI, and others. We are actively expanding our partner network.
What does an inference partner provide?
Inference partners contribute GPU compute to host open-source models on the gateway. Partners retain visibility for their participation and help demonstrate that public, open-source AI can be delivered outside the orbit of big tech.
Can inference partners make their own announcements?
Yes. We encourage partners to share their participation. We ask that partner announcements align with central Public AI messaging. A comms toolkit with sample language can be provided upon request.
How does coordination work?
- Dedicated channels are set up for technical and communications coordination
- Each partner should designate a point of contact for technical and communications matters
Getting Involved
Ready to become an inference partner? Contact us at [email protected] with:
- Your organization details
- Available compute resources
- Technical contact information
- Preferred communication channels
We'll provide you with the technical integration guide and partnership agreement.