Langtrace is an open source observability and evaluations platform for AI agents. It provides tools to track token usage, cost, latency, and accuracy, explore API requests with traces, run evaluations, version-control prompts, and compare prompt performance across models. Setup is non-intrusive with two lines of code in Python or TypeScript. Langtrace supports popular frameworks including CrewAI, DSPy, LlamaIndex, and Langchain, as well as various LLM providers and vector databases. It offers SOC2 Type II compliance and enterprise-grade security. The platform is fully open source with integrations for over 30 tools.
Key Benefits
- Simple non-intrusive setup with 2 lines of code
- Open source with full access to source code on GitHub
- SOC2 Type II compliant for enterprise security
- Supports multiple frameworks and LLM providers