telemetry.dev
by @ephraimduncanMonitoring and Observability for AI and LLM apps — trace every model call, tool step, and retrieval with tokens, cost, latency, and errors in one place.
Telemetry is an OpenTelemetry-native observability platform built specifically for AI and LLM applications. It captures every model call, tool step, and retrieval in a single unified trace, attaching tokens, cost, latency, and error data automatically. Developers can send their first trace in under five minutes using drop-in SDKs or standard OTLP ingest, with no framework or language lock-in. Key features include real-time failure streaming, input/output capture with secret redaction, cost and token trend analytics, and flexible grouping by model, provider, or environment. A free tier includes 10,000 ingestion units per month with no credit card required, making it accessible for teams at any scale.