All capabilities
Connect
Multi-model routing
Supply the provider and model on create without tying your runtime to one provider.
- The caller supplies the provider and model on create.
Provider and model
The caller supplies the provider and model on create.
- 01
Supply the provider and model on create
- 02
Let a runtime request the configured model
- 03
Inspect model usage alongside agent activity
A practical use case
Route each request to a configured model while the runtime remains responsible for inference.
- Multiple model providers
- Runtime-selected inference
- Model access without product-side provider coupling
Capabilities
Provider and model
The caller supplies the provider and model on create.
The caller supplies the provider and model on create.
The caller supplies the provider and model on create.
Start integrating
Start integrating
Read the integration guide, then connect your runtime and backend.