AI Model Inference
Hosted inference for large language models and other AI workloads, billed per token or per request.
What each cloud calls it
US hyperscalers
- AWSUSBedrock
- AzureUSAzure OpenAI Service
- GCPUSVertex AI
European clouds
- ScalewayFRGenerative APIs
- OVHcloudFRAI Endpoints
- StackITDEAI Model Serving
- IONOSDEAI Model Hub
- HetznerDENo direct equivalent today
Want to see how this maps to your full stack? Back to the full comparison table.