All Tools
L
MonitoringFreemiumOpen Source
LANGTRACE
Open-source LLM observability with OpenTelemetry
AGPL-3.0
ABOUT
LLM applications are notoriously hard to debug and monitor. Traditional APM tools don't understand LLM-specific concepts like prompts, completions, token usage, vector DB queries, or agent tool calls. Langtrace provides OpenTelemetry-native instrumentation that captures LLM calls, framework operations, and vector database queries in a single dashboard with real-time tracing, cost tracking, latency analysis, and evaluation metrics.
INSTALL
pip install langtrace-python-sdkINTEGRATION GUIDE
1. Monitor real-time LLM API calls across OpenAI, Anthropic, and other providers with detailed tracing and cost breakdowns
2. Debug agentic workflows by inspecting tool calls, LLM responses, and vector database queries in a unified trace view
3. Run automated evaluations on LLM outputs to detect regressions after prompt or model changes
4. Track token usage and latency trends over time to optimize model selection and prompt design
5. Self-host the platform for compliance-sensitive environments requiring data sovereignty
TAGS
observabilitytracingopentelemetryllmmonitoringevaluationpythontypescript