Developer preview
Connect your inference traffic
Sign in to create a workspace and API key. Then initialize the SDK once; your existing model calls stay unchanged.
Already have an API key?
Overview
Inference traffic
Connecting
Calls—
Operations—
Providers—
Average latency—
Errors—
Monitoring
Observed operations
Calls are grouped by provider, method, model, and application callsite.
| Operation | Provider | Model | Calls | Success | Avg. latency | Status |
|---|
No inference calls observed yet
Initialize the SDK, run your application, and calls from supported clients will appear here.
import { init } from "@smallermodels/node";
init({ apiKey: process.env.SMALLERMODELS_API_KEY });Monitoring
Recent calls
Metadata only. Request and response content is not captured by default.
| Time | Provider | Method | Model | Latency | Result |
|---|
No calls recorded yet.
Workspace settings
API keys
Create and revoke credentials used by the SDK. Full key values are shown once.
New API key
Store this value now. It cannot be shown again.
No API keys yet.