Sending Data In
Observability starts with getting your AI activity into the platform. There are three ways to send traces — pick whichever fits your stack. All of them authenticate with your project API keys and produce the same traces and observations.
Create project API keys under Settings → API Keys (see API Keys), then configure your SDK or exporter with them.
1. SDKs
The native SDKs give you the richest instrumentation — manual tracing, decorators, and auto-instrumentation of popular LLM clients.
2. OpenTelemetry (OTLP)
If you already emit OpenTelemetry, point your OTLP exporter at the platform's OTLP endpoint and your spans become traces — including GenAI semantic conventions, agent frameworks, and coding agents.
3. Framework integrations
Drop-in integrations instrument popular frameworks with a few lines:
What to send
To get the most out of observability, include:
sessionId— to group traces into sessions.userId— to enable per-user analytics.tags,metadata, andenvironment— to filter and slice traces.- Model, token, and cost details on generations — for accurate FinOps. Costs are computed from your model definitions.
Next steps
- Traces & Observations — explore what you sent.
- Integrations — the full list of SDKs and connectors.
- API Reference — the ingestion endpoint and REST API.