Traceloop
Open-source LLM observability built on OpenTelemetry, with automatic instrumentation for major providers and frameworks
Traceloop provides LLM observability through OpenLLMetry, an open-source set of OpenTelemetry extensions for GenAI applications. It automatically captures latency, token usage, costs, and full prompt/response pairs with just two lines of code. Supports OpenAI, Anthropic, Cohere, AWS Bedrock, and Google Vertex AI, plus frameworks like LangChain, LlamaIndex, and Haystack. SDKs available for Python, TypeScript, Go, and Ruby. Because it is OpenTelemetry-native, traces can be exported to any compatible backend including Datadog, New Relic, Honeycomb, and Sentry.
Pricing: Monthly subscriptions
Traceloop Alternatives
Explore 45 products in the Observability & Analytics category. View all Traceloop alternatives.
RAGAS
Open-source evaluation and testing framework for LLM and RAG applications
Guardrails AI
Open-source framework for adding input and output validators around LLM calls
NeMo Guardrails
NVIDIA toolkit for adding programmable guardrails to LLM conversational apps
Presidio
Microsoft open-source SDK for detecting and anonymizing PII in text and images
Work on Traceloop? Feature it at the top of Observability & Analytics.
Is your product missing?